Title: Behavioral effects of dextran-induced intravascular aggregation of red blood cells (sludge) in the rabbit
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00097848/00001
 Material Information
Title: Behavioral effects of dextran-induced intravascular aggregation of red blood cells (sludge) in the rabbit
Physical Description: ix, 171 leaves. : illus. ; 28 cm.
Language: English
Creator: Brooks, Karl Michael, 1926-
Publication Date: 1966
Copyright Date: 1966
Subject: Blood -- Agglutination   ( lcsh )
Behaviorism (Psychology)   ( lcsh )
Rabbits   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis -- University of Florida, 196.
Bibliography: Bibliography: leaves 159-170.
Additional Physical Form: Also available on World Wide Web
General Note: Manuscript copy.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097848
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000543417
oclc - 13153338
notis - ACW7125


This item has the following downloads:

PDF ( 7 MBs ) ( PDF )

Full Text







April, 1966


This work was supported by the Agricultural Experiment Station

and was carried out in the laboratories of the Department of Food

Technology and Nutrition. The author wishes to express his deep appre-

ciation to Dr. Raymond A. Dennison and to the staff of the Department

of Food Technology and Nutrition for their friendly interest and


The author also wishes to express his gratitude to Dr. Bradford N.

Bunnell for supervising the conduct of the research and to Dr. Wilse B.

Webb, Dr. Frederick A. King and Dr. Robert L. King for their advice and

counsel as members of the supervisory committee. Special thanks are

extended to Dr. Ralph C. Robbins who not only introduced the author to

the problem area but who as a member of the supervisory committee served

as a continuing source of encouragement.

The provocative criticisms and advice of Dr. Rowland B. French

and the enthusiastic assistance of Mr. Howard L. Povey are gratefully


Finally, the author wishes to acknowledge his debt to his wife,

Sally, for her very capable assistance in gathering and analyzing the

data and for her typing of the manuscript. No criticism is implied in

noting that the typing was interrupted by the sudden arrival of a first

born baby girl named Julie Elizabeth.



ACKNOLEDG ETS . . . . . . . . . . ii

LIST OF TABLES . . . . . . . . ... . . . v

LIST OF FIGURES . . . . . . . . ... .... viii

INTRODUCTION ..... .. ..... .............. 1

Background . . . . . . . . ... . .. 1

History of Sludge . . . . . . . . . 2

Function of Sludge . . . . . .. . . . 5

Sludge as an Indicator of Pathology . . ... .
The Effects of'Sludge . . . . . . . 8

Present Status of the Sludge Problem . . . .. .30


Part I. Preliminary Problems . . . . . .. .33

Appropriate Subject . . . . . . ... .33
A Method for Inducing the Sludge . . . . 35
Evaluation of Level of Sludge . . . . .. 40
Behavioral Tests . . . . . . . .

Part II. Measuring the Blood Sludge . . . ... .45

Experiment 1 . . . . . . . ...
Experiment 2 . . . . . . . . ... 51

Part III. Behavioral Studies . . . . . .. .53

Experiment 1 . . . . . . . . .
Experiment 2 . . . . . . . . . 73
Experiment 3 . . . . . . . . ... 81
Experiment 4 . . . . . . . .... .91

GENERAL DISCUSSION ...................... 106


SUaLiRY . . . . . . . . . . . . . . 117

APPENDICES ........................ .. 120

Appen..x A ......... ... ............ 121

Appendix B ........ .... .... .. ..... .130

Appendix C .. . . ... . . . . . .. 141

Appendix D ........... ....... ... . 14l

Appendix E .................. .... 150

REFERE CES ............... ........ . 159

BICGRAPHICAL SKETCH .................... . 171


Table Page

1. Mean Daily Food Intake (gms) During 2 Days Before and for
4 Days After Treatments with Five Animals per Group . . 62

2. Mean Daily Water Intake (gms) During 2 Days Before and for
h Days After Treatments with Five Animals per Group . .. 63

3. Mean Body Weight (gms) During 2 Days Before and for 4 Days
After Treatments with Five Animals per Group . . . 65

l. Mean Home-cage Activity (pacing) During 2 Days Before and
for 4 Days After Treatments with Five Animals per Group . 66

5. Mean Exploratory Behavior During 2 Days Before and for h
Days After the Treatments with Five Animals per Group . 68

6. Mean Blood Sludge Levels During 2 Days Before and for 4
Days After Treatments with Five Animals per Group . . 68

7. Mean Response Latencies (in sec.) for First Trials from
Three Successive Series of Three Trials Each for Four
Groups of Albino Rabbits in a Shock-escape Situation with
Six Animals per Group . . . . . . . . . 76

8. Mean Shock-escape Latencies (in sec.) for Combined Scores for
Three Successive Series of Learning Trials with Six Ani-
mals per Group . . . . . . . .... . .. 77

9. Group M1ean Blood Smear Ratings Before (Smear A) and After
(Smear B) Treatments and After Shock-escape Conditioning
Trials (Smear C) with Six Animals per Group ....... 79

10. Mean Response Latencies and Running Times (in sec.) for
First trialss From Each of Three 5-Trial Food Reinforced
Test Series and One Extinction Series (4) for Four Groups
of Rabbits with Six Animals per Group . . .... .. 85

11. Mean Total Response Latencies and Running Times (in sec.)
for Each of Three Successive 5-Trial Series of Food Re-
inforced Learning Trials and One Extinction Series (4)
for Four Groups of Six Albino Rabbits Tested in a h-Com-
partment "Square Maze" . . . . . . .. . ... 86

Table Page

12. Mean Blood Sludge Levels Before (Smear A) and After
Treatments (Smear B) and After Four Series of Training
Trials (C) in the "Square Maze" with Six Animals per
Group . . . . . . . . . . . . . 88

13. Mean Performance Measures for All Rabbits from Experi-
ment h Shown Separately for Each Treatment Condition . 96


12. Food Intake (gms) per 12 Hour Interval and par Day for
Each Aninal and Group with Treatments Given at the End of
Dny 2 . . . . . . . . . . . .. .. 130

15. WYater Intake (gms) per 12 Hour Interval and per Day for
Each Animal and Group with Treatments Given at the End
of Day 2 . . . . . . . ... .. .. . 132

16. Mean Daily Body Weights (gms) for All Animals During 6
Days of Observations Made at 12 Hour Intervals with
Treatments Given at the End of Day 2 . . . . . 13h

17. Home-cage Activity (number of round trips) for Each
Animal and Groun for Each 12 Hour Interval Beginning 2
Days Before Treatments . . . . . . . . 136

18. Exploratory Activity (compartments entered) by Four
Groups of Albino Rabbits During 12 Successive 1 Minute
Ixposures to a 4-Compartment "Square M.aze" Over the 6 Day
Observation Period Beginning at 2 Days Before Treatment 138

19. Blood Sludge Ratings, Based on Scale Described in Exper-
iment 2, Part II, for Blood Smears Obtained From HmDx-
Treated Animals at 12 Hour Intervals Beginning at 2 Days
Before Treatments . . . . . . . .. . . 110

20. Shock-escape Latencies (in sec.) for Three Successive 3-
Trial Series for Six Replications Using Rabbits Tested
at 1 Hour After Treatment .. . . . . . . 121

21. Blood Sludge Ratings for Blood Smears Taken Before (A)
and After (B) Treatments and After Shock-escape Condi-
tioning Trials (C) . . . . . . . .... .43

22. Response Latencies and Running Times (in sec.) for First
Trials From -ach of Three 5-Trial Food Reinforced Test
Series and One Extinction Series ()) Using Albino Rabbits
Under Four Different Experimental Conditions ..... .. .1

Table Page

23. Total Response Latencies and Running Times (in sec.) for
Three Successive Sets of Food Reinforced Learning Trials
and One Extinction Series (l) in Four Groups of Albino
Rabbits Tested in a L-Compartment "Square Maze" ...... 16

2h. Food Intake (Gms) During a 2 Hour Feeding Period Follow-
ing Food Reinforced Learning Trials . . . . ..1h8

25. Blood Sludge Ratings for Blood Smears Taken Before (A) and
After (B) Treatments ani After Food Reinforced Learning
Trials (C) for Four Groups of Rabbits. . . . .. 119

26. Total Responses (T) and Reinforced Responses (R) Before
(A) and After (B) Treatment and During Extinction Series
(C) 2h Hours After Treatments for Eight Rabbits Repre-
senting Two Replications of a h x L Latin Square .... 150

27. Analysis of Variance for the Data Obtained From Exper-
iment h. Separate Analyses are Shown for Total Responses,
Reinforced Responses, Per Cent Reinforced Responses, and
Number of Responses per Reinforcement. Differences Scores
1Tere Used in Every Instance . . . . . . . 152

28. Blood Sludge Measures From Eight Rabbits Representing
Two Latin Square Replications of a Test of Bar Press Per-
formance on a DRL Schedule. Smears Were Taken Before
(Smear A) and After (Smear B) Treatment, After the 1
Hour Test Session (Smear C), and 2h Hours Later Follow-
ing the 1 Hour Extinction Series (Smear D). . . . 156

29. Food Intake (Gms) by Albino Rabbits During Each 2 Hour
Posttreatment Feeding Period Follovring Testing on a
Bar Press Task Using a DRL Reinforcement Schedule . . 157

30. Performance on a Bar Press Task Under a DRL Schedule by
Albino Rabbits During 5 Days Before and for 5 Days After
Treatments . . . . . . . . ... . . . 158


Figure Page

1. Mean Food Intake per 21 Hours for Each of Four Groups
of Albino Rabbits with Five Animals per Group . . . 6

2. Mean rater Intake per 2h Hours for Each of Four Groups
of Rabbits with Five Animals per Group . . . . 6h

3. Mean 2h Hour Body 7Yeights for Each of Four Groups of
Five Albino Rabbits During 6 Days of Observation with
Treatment Given at the End of 2 Days . . . . . 67

L. Mean Home-cage Activity (No. Round Trips) for Four
Groups of Five Albino Rabbits Before and After Treat-
ments During the 6 Day Observation Period . . .... 67

5. Mean number of "Square Maze" Compartments Entered by Four
Groups of Five Albino Rabbits Tested Daily at 12 Hour
Intervals Over the 6'Day Period Beginning 2 Days Before
Treatments. . . . . . . . . 69

6. Mean Blood Sludge Levels During 2 Days Before and for L
Days After Treatments for Four Groups of Five Rabbits . 69

7. Mean Response Latencies for First Trials for Successive
Series of Shock-escape Learning Trials . . . .. 78

8. Mean Total Response Latencies (in Sec.) for Four Groups
of Six Albino Rabbits for Each of Three Successive Series
of Three Trials Each in a Shock-escape Learning Situa-
tion . . . . . . . . ... . . .... 78

9. Mean Groups Results Using First Trials for Measures of
Response Latency and Running Time from Four Groups of
Six Albino Rabbits Receiving Three Series of Food Re-
inforced Training Trials and One Extinction Series . 87

10. Mean Group Results for Measures of Response Latency and
Running Time Obtained from Four Groups of Six Albino
Rabbits Tested in a "Square Maze" with Three Series of
Positively Reinforced Learning Trials and One Extinction
Series . . . . . . . . ... . ... .87


Figure Page

11. Mean Performance for All Animals Before and After Each
of Four Separate Treatment Conditions. Results Shown
Include Mean Total Responses, TMean Reinforced Responses,
Mean Per Cent Reinforced Responses and Mean Number of
Responses per Reinforcement. . . . . . . ... 97

12. Comparison of Blood Smears and In Vivo Observations of the
Condition of the Blood in the Mesentery of the Rabbit
Before and After Treatment with Standard Dose Levels of
Saline, 77k Dextran or HmDx . . . . . . . 121

13. Condition of the Blood Smear Before Treatment and After
Treatment with 20, 50 and 100 Per Cent of Normal Dose
of HmDx. Each Set of Four Photographs Includes Two
Observations From Separate (A and B) Locations From
Each of Two Blood Smears (1 and 2) Taken at the
Same Time. The Four Sets of Photographs Were Used
as Standards to Represent Blood Sludge Levels +1
Throu h +h, Respectively, for Rating the Blood
Smears Obtained During the Course of the Behavior
Studies . . . . . . . . .. . . . 12


A series of experiments was carried out to investigate the effect

of severe intravascular aggregation of red blood cells ("sludge") on

behavior. Observations were made on levels of food and water intake,

on home-cage activity and exploratory behavior, on response latencies

during successive exposures to a shock-escape learning situation, on

response latency and running time during successive exposures to a food

reinforced maze-runway learning problem and on bar press performance

under a DRL (differential reinforcement of low response rates) schedule.

The sludging was induced by the infusion of 1 gm/kg body weight of

2000k molecular weight dextran and evaluated by means of blood smears.

Separate groups were used to control for the effects of the dextran per

se and of the saline solution.

No statistically significant differences attributable to the

effects of the dextran-induced sludge were found on any of the behavior-

al measures.


The phenomenon of intravascular aggregation as a concomitant

to illness and injury has been known under a variety of names and guises

since the time of Hippocrates (Bigelow, Heimbecker and Harrison, 19h9;

Fahraeus, 1929; Knisely, 1951; Thygesen, 19L2).

It has been demonstrated in a large number of disease states


(Ditzel, 1955; Fahraeus, 1929; Hirschboek and Woo, 1950; Knisely and

Bloch, 1942) and in pregnancy (Fahraeus, 1929; Odell, Aragon and Pottin-

ger, 1947). It has been produced by the traumatic manipulation of

tissue such as burning (Brooks, Dragstedt, Warner and Knisely, 1950;

Gelin, 1959), crushing (Gelin, 1956; Knisely, Eliot and Bloch, 1955),

by hypothermia (Lofstrom, 1959), by feeding a high fat diet (Swank and

Cullen, 1953), by the infusion of high molecular weight substances such

as thrombin, fibrinogen, gelatin or dextran (Cullen and Swank, 195h;

Thorsen and Hint, 1950) and even by ingestion of alcohol (Bloch, 1956;

Knisely, 1951).

A variety of processes might contribute to the aggregation under

the several conditions listed above (Gelin, 1956; Gelin, 1959; Thyge-

sen, 1942). Gelin (1959) has classified these alternatives as specific

(agglutination or coagulation) and non-specific (pseudo-agglutination)

factors. The specific processes include, for example, the antigen-

antibody reaction and the conversion of fibrinogen into fibrin. The

non-specific processes relate to the aggregation generally associated

with the increase in the proportion of high molecular weight substances

in the blood stream and to the clumping which has been found to occur

with the complete interruption of blood flow (Gelin, 1956).

It should be emphasized that the present paper will be concerned

with red cell aggregation induced by non-specific processes (pseudo-

agglutination), although the pathologic effects of the aggregation may

be independent of the process by which the aggregation was induced.

History of Sludge

The earliest reported observations of experimentally produced


intravascular aggregation were those of Joseph Lister (1858) who as

early as 1827 (Norris, 1869) studied changes in the blood flowing

through the small vessels in the wing of the bat following injury to

the tissue. Richard Norris (1862, 1869) not only made extensive obser-

vation of experimentally produced red cell aggregation in the web of

the frog, but studied the mechanism of aggregation by adding such sub-

stances as gum, gelatin and very thick solutions of starch to samples

of blood. By adding increasing amounts of these substances he saw the

character of the cellular aggregates change from that of mild rouleaux

to firm masses of cells in which the outlines of the cells became less

and less distinct until the individual identities of the cells were

wholly lost.

Norris (1869) also described the process of venous stasis and the

development of inflammation through the loss of fluid from the dilated

vessels. This effect was usually associated with a condition of homo-

geneous stasis in which the outlines of the individual red cells could

not be recognized.

Red cell aggregation was given a name ("sludge") and widely pub-

licized through the work of Melvin Knisely and his coworkers (Knisely,

1951). Knisely (1936, 1938) had developed a fused quartz rod illumi-

nator with which he was able to make continuous observations of living

tissue without truamatizing the tissue itself. He began by studying

the vascular system of the normal animal (Knisely, 1940) and eventually

applied his techniques to the study of blood flow in pathological states.

He first studied malaria in the monkey. The monkey was experi-

mentally infected with Plasmodium Knowlesi malaria. Blood flow in the

internal organs was studied and photographed throughout the course of


the disease. The motion pictures taken of the severe aggregation during

Stage III of the Knowlesi malaria were shown publicly (Knisely, Stratman-

Thomas and Eliot, 1940). There followed a number of papers and motion

pictures dealing with red cell aggregation (which Knisely and his co-

workers called "sludge") in the experimentally infected malarial monkey

(Knisely, Stratman-Thomas and Eliot, 1911) and in canaries experimentally

infected with avian malaria (Lack, 1912). Blood flow and "sludge" were

also observed in a large number of human disease states (Knisely and

Bloch, 1942).

In 1914, Knisely and his coworkers produced potentially lethal

sludging in the monkey by infecting it with malaria and then reversed

the condition with atabrine or quinine. Later these same researchers

(Knisely, Stratman-Thomas, Eliot and Bloch, 1919, 1951) gave the malaria

infected monkeys heparin shortly before the parasites sengented. Under

these conditions the severe sludging did not occur although the para-

sites in the blood multiplied greatly. Those animals which showed severe

sludging died regardless of parasite count while the ones receiving the

heparin showed high parasite counts but few outward signs of the illness.

It was also noted that those animals developing the thick pasty sludge

went into "slowly deepening comatose condition ending in a deep coma

before death 't (Knisely, Bloch, Eliot and Varner, 1917).

In 1947, Knisely and his coworkers published a paper entitled

"Sludged Blood" in which they described various types of sludge and

reviewed and extended the arguments for the pathologic significance of

sludge. Sludge was supposed to have two functions: (1) it was a patho-

logic indicator; (2) it had of itself a highly significant effect on the

course of disease.


Functions of Sludge

Sludge as an indicator of pathology

According to the investigators referred to above, sludge was a

general term for aggregates of many different sizes and kinds. There

were "basic masses" aggregates which were sufficiently firm or rigid

to resist breakup upon passage through the peripheral vessels; such

basics might vary greatly in size. Depending on the severity of the

aggregation, some or all of the red cells might be clumped. The several

types of sludge described by Knisely et al. had one thing in common.

There were always some rigid masses.

A definition of healthy, normal blood was provided as a baseline

against which to measure pathology. In all organisms completely healthy

blood flow is "streamlined", or laminar, and always so rapid that in

vessels of from 60 to 120 microns it is impossible to distinguish the

individual red cells. When observed in vitro, healthy red cells have

sharply defined boundaries and the cells tend to repel each other within

the blood fluid. As this blood flows through the capillaries, there is

a minimum of loss of fluid through the capillaries and no detectable


Knisely's speculations concerning sludge were an invitation to

controversy. First of all, Fahraeus's classic review (1929) had equated

red cell aggregation with rouleaux and mild rouleaux with normal good

health. Knisely (Knisely, Eliot and Bloch, 1965) stated emphatically

that what he was studying was "not rouleaux", but masses of cells within

which the individual cell lost its shape and identity. Apparently nei-

ther author was familiar with Richard Norris' (1869) demonstration that


rouleaux and the formless masses of cells represented only different

degrees of the same phenomenon. Further, had either author been famil-

iar with the papers by Lister (1858) or Norris (1862, 1869), the issue

of whether mild sludging was compatible with good health might never

have arisen since the earlier writers had pointed out that blood which

gave no indication of aggregation in vivo, even during complete stasis,

nevertheless readily formed rouleaux when placed on slides. And Knisely

and his coworkers made their observations in vivo while Fahraeus used

in vitro techniques almost exclusively.

The significance of sludge as an indicator of pathology remained

a lively issue. Lutz, Fulton and Akers (1951) reported that they were

unable to demonstrate rigid cellular aggregates in the hamster follow-

ing trauma, burns or neoplasia. Also, a number of investigators, basing

their arguments largely on clinical studies, reported that sludge in the

conjunctival vessels was a common finding even in the absence of clini-

cal evidence of pathology (Lack, Adolph, Ralston, Leiby, Winsor and

Griffith, 19h9; Robertson, Wolf and Wolff, 1950), and that sludge was of

little diagnostic significance.

Robertson, Wolf and Wolff (1950) reported that they could find no

correlation between sludge and clinical symptoms of illness. Further,

they stated that sludging even within the individual patient varied

greatly from day to day and from blood vessel to blood vessel. To them

flow rate seemed to be the critical variable since presence of sludge

was found to correlate only with degree of vasodilation. It was there-

fore hypothesized that the observed sludging was the result rather than

the cause of the reduced blood flow rate. To test this hypothesis vaso-

constriction was induced in one eye while the other eye served as control.


Following application of neosynephrine solution there was severe vaso-

constriction and the disappearance of sludge from the blood stream while

the control eye continued to show sludging. Then, when vasodilation was

induced by histamine phosphate, sludge appeared where before there had

been a complete absence of sludge. It was suggested that blood flow in

the conjunctivae is the result of an increase in the volume of the local

vascular beds and a decrease in the velocity of blood flow. Moreover,

these experimenters were confident that "widespread sludging is compati-

ble with good health and well being".

The temporary sludging that Fowler (1949) saw in conjunctival

vessels following cervical sympathetic nervous stimulation or injections

of noradrenaline may have been related to flow changes alone especially

since small doses of novacaine reversed the effect. However, Lutz

(1951) offered the alternative suggestion that this "sludge" may have

been merely an illusion produced by rhythmic contractions of the small

vessels of the conjunctiva.

All of the above findings were consistent with the implications

of Ploman's (1920) demonstration that red cell aggregation could be pro-

duced in vivo, at least in man, by pressing the eyeball and inducing sta-

sis within the retinal vessels. The resulting aggregation, or reaggre-

gation, could then be observed in the larger venules and veins by opthal-

moscopy. Ploman was a student of Fahraeus (1929); and his experiment

contributed to Fahraeus' conviction that rouleaux represented a cou-

pletely normal condition of the blood.

Laufman (1951) reviewed some of the evidence against Knisely's

arguments and conceded the possibility that sludge may be such a sensi-

tive indicator of pathology that the ailment may be undetectable by


ordinary clinical tests even though sludge is present. This was what

Knisely and his coworkers believed (Knisely et al., 19h7). They had

studied 50 normal, healthy medical students and student nurses and found

an absence of sludge. Fahraeus (1929) had stated that the healthy horse

shows a high red cell sedimentation rate and a strong tendency to red

cell aggregation, an observation which he considered as strong evidence

for his conclusion that red cell clumping is fully compatible with good

health. In an attack on Fahraeus' argument, Knisely, Bloch, Brooks and

?'arner (1950) made their own observations (they looked into the horse's

eye) and found that sludging in the healthy horse is associated with

fright and rough handling. Sludge, it seemed, was not a normal condi-

tion even in the horse. On the other hand, hundreds of seriously ill

patients representing a wide range of diagnoses, including some psy-

chiatric cases, which were studied by Knisely and his coworkers (19h7),

all showed sludge ("not rouleaux"). Thus, according to these workers,

sludge does not occur in health, but it does occur in pathology.

In spite of Knisely's confidence, the question of whether sludge

represents a sensitive indicator of pathology remains unresolved. Pos-

sibly the resolution of this issue must wait for a more reliable means

for measuring the sludge.

The effects of sludge

Sludge and the body. Knisely and his coworkers (Knisely, Bloch,

Eliot and 'Warner, 1950) also remained convinced that sludge had a high-

ly significant and detrimental effect upon the organism. Sludge, they

were sure, could damage the body through a number of mechanisms. The


blood flow rate is reduced. This in turn restricts the transport of

oxygen and glucose to the tissues including the vascular endothelium.

Stagnant anoxia of the tissues develops. The anoxic walls of the capil-

laries become increasingly permeable to the plasma proteins and increas-

ing quantities of proteins and blood fluid are lost into the surrounding

tissues. Impacted cells and fluid loss reduce blood volume. Clumped

red cells are destroyed by the phagocyte cells of the liver and spleen,

promoting a condition of anemia.

Histological attempts to implicate tissue anoxia in death follow-

ing severe sludging failed; there was no indication of the cause of

death (Knisely, Stratman-Thomas, Eliot and Bloch, 1951). It was argued,

however, that if tissue anoxia were the cause of death, then comparison

with the "normal" should yield no new information since "... all, or

almost all, of the tissues studied by the student of "normal histology"

have died or been killed by processes directly involving anoxia"

(Knisely, 1951, p.88).

In the "thick, mucklike sludge" such as Knisely et al. (19L5)

found in the terminal stage of Knowlesi malaria in the monkey there may

be justification for the tissue anoxia hypothesis in spite of lack of

histological verification; but for less severe sludge perhaps the hypo-

thesis should sound more like a question. Landis (1928) did show that

stagnant anoxia produced by compressing the mesenteric artery and vein

produced in about 3'min. an increase in the permeability of the capil-

lary walls sufficient to increase the rate of fluid loss to four times

above normal, and to permit the loss of such quantities of plasma pro-

teins that the osmotic pressure of the remaining blood was reduced to

one-half of its previous value. It should be noted, however, that


immediately upon renewal of circulation the capillary permeability

returned to normal. With less than complete anoxia the effect upon the

blood vessel walls appears to be extremely difficult to demonstrate.

Also, Norris (1869) observed spontaneous deaggregation and a sudden and

apparently complete recovery of normal vascular function after a 3

hour period of carplete capillary stasis in the frog. Perfusion studies

suggest that the vascular endothelium is actually highly resistant to

anoxia since the oxygen tension of perfusion fluids must be reduced al-

most to zero before the endothelium is affected (Zweifach, 1961).

Finally Van Liere and Stickney (1963) after reviewing the literature

relating to endothelial hypoxia concluded that hypoxia within the phy-

siological range has little effect if any on capilliary permeability.

Muscle tissue also seems to be highly resistant to oxygen lack.

Stainsby and Otis (1964) found that oxygen uptake by resting muscle was

not affected by reduction of the blood oxygen tension until a critical

low value occurred. Further, the contracting muscle with eight times

the oxygen consumption of the resting muscle was found to have a lower

oxygen tension than the resting muscle. Changes in capillary density

were advanced to explain these results. This explanation is consistent

with Krogh's (1919) estimate that the number of open capillaries, which

may be less than 100 per mm3 in the resting muscle, may rise to more

than 2500 per mm3 when the muscle is activated.

Yet it is conceivable that even this reserve capacity of the cap-

illary networks may not be adequate at all times. Venous blood from

traumatized tissue has been found to have a high oxygen tension

(Blalock, 1943) even though severe sludging occurs at the site of any

tissue trauma (Knisely, Eliot and Bloch, 1945; Gelin, 1956). However,


since arteriovenous shunts consistently dilate in response to blockage

of capillary beds (Heimbecher and Bigelow, 1950) this increase in oxygen

content of the venous blood may be associated with reduced capillary


Sludging and vascular stasis have been shown to be associated

with inflammation and tissue trauma (Knisely, Bloch, Brooks and Warner,

1950; Lister, 1858; Norris, 1869). Yet Gessler (1932) found a 10 fold

increase in metabolic rate in the inflamed tissue of the rabbit's ear.

.hile such highly localized metabolic changes may be explained, at least

in part, by increases in the rate of anaerobic glycolysis (Frunder,

1953), the finding of sustained increases in oxygen consumption of 30 to

60 per cent in severely burned patients during the early postburn period

(Cope, Nardi, Quijano, Rovit, Stanbury and Wight, 1953), when the sludg-

ing should already have been severe (Brooks, Dragstedt, Warner and

Knisely, 1950; Gelin, 1956), seems to present no alternative to the con-

clusion that sludge may be associated with an increase in 02 uptake.

Even scorbutic guinea pigs show a level of 02 uptake 20 per cent higher

than normals (Evans and Hughes, 1964) in spite of the fact that the con-

dition is associated with severe blood sludge (Robbins, to be published).

However, Lofstrom (1959), studying the effect of sludging on oxy-

gen uptake, induced the sludge with infusion of high molecular dextran

and subsequently reversed the effect with low molecular weight dextran.

He found that oxygen uptake decreased significantly with the sludging

and recovered as blood flow improved. While the obvious conclusion from

this study has been widely accepted (Gelin and Zederfeldt, 1961; Long,

Sanchez, Varco and Lillehei, 1961), attempts by the present author to

replicate these findings have failed (Robbins, 1963-1964).


Lofstrom (1959) was also able to show what appeared to be an

oxygen deficit during the rearming phase of hypothermia. Infusions of

low molecular weight dextran reduced sludge and improved flow rate and

brought an increase in oxygen uptake. He suggested that scme tissues do

develop serious oxygen deficits and even acidosis during severe sludge.

Such a differential effect should then be difficult to detect through

analysis of the peripheral blood because of the delayed removal of the

acid metabolites into the blood stream.

Gelin (1956) found visceral damage following a prolonged period

of severe sludging and recommended "tissue anoxia" as the causal mechan-

ism. The sludging was produced by fracture, contusions, burns, and by

infusions of thrombin or high molecular weight dextran. All treatments

resulted in profound sludging. All animals were sacrificed at the end of

the third day. Inspection of visceral organs revealed necrotic areas in

the livers, degenerative changes in the kidney tubules and hemmorhages

in the lungs; and the pathology appeared to be largely independent of

the means by which the intravascular aggregation was induced. Apparent-

ly the anoxiaa" hypothesis was justified both because it followed logi-

cally from the observation that during severe sludging blood flow rate

is much reduced and because of the nature of the resulting injuries.

Severe sludging is a common finding during and following extra-

corporeal circulation (Clowes, 1960). Tissue damage such as microin-

farctions (much like that described by Gelin, 1956) of the kidney, liver,

and myocardium are also a common finding in prolonged total body perfu-

sion. However, such damage has been prevented through the use of low

molecular weight dextran, which prevents the aggregation (Finsterbusch,

Long, Sellers, Amplatz and Lillehei, 1961).


While the mechanism for the selective tissue damage associated

with the sludge has not been fully demonstrated, the evidence seems to

be consistent with the view that selective redistribution of the blood

flow by sympathetic nervous activity compounds the effects of the sludge

to produce localized tissue anoxia and acidosis (Gelin, 1962; Long, San-

chez, Varco and Lillehei, 1961). Gelin (1962) has noted that studies of

the metabolic effects of induced red cell aggregation suggest the accum-

ulation of acid materials in the tissues during aggregation with their

subsequent removal when flow is improved. This hypothesis is based on

the finding of a transient acidosis together with a reduced blood pH and

a rise in pCO2 and lactic acid which follows the increase in blood flow.

Additional evidence indicates, however, that the mechanism under-

lying the pathologic action of severe intravascular aggregation may be

related to abnormal increases in oxygen uptake as well as to reduced

oxygen availability. Gilmore and Fozzard (1960) studied liver function

following severe thermal trauma to 30 percent of the body, a condition

which should have resulted in very severe sludging (Brooks et al., 1950).

Under these conditions the arteriovenous 02 difference rose and there

was an increase in oxygen uptake in the liver. It was their conclusion

that "... hepatic hypoxia does not contribute to the production of early

hepatic injury following severe thermal trauma.? On the other hand,

Hinshaw, Pories, Harris, Davis and Schwartz (1960) have reported a reduc-

ed oxygen tension in liver and kidneys following single injections of

high molecular weight dextran in the anesthetized dog. The differences

between these two sets of observations could reflect a difference in

levels of sludge, and presumably therefore also in blood flow rates.

The evidence appears to be generally favorable to the idea that


sludge promotes a condition of stagnant anoxia and that this effect

represents a mediating mechanism for the pathophysiologic consequences

of sludge.

At this point all of the critical variables seem to be joined.

Stagnant anoxia mediates the effects of sludge. The nervous system me-

diates behavior. The highly active neural tissue should be sensitive to

the effects of anoxia and so it should also be sensitive to the sludge.

Sludge and the nervous system. Knisely and his coworkers felt

that the nervous system was particularly vulnerable to the effects of

sludge. Upon finding sludge in the eye of a psychiatric patient they

hypothesized that the patient's nervous system had been permanently

damaged by capillary plugs of red cells masses; and the assumed mecha-

nism of action of the sludge became evident "when one considers the par-

allelism between the known effects on normal persons of breathing slowly

decreasing concentrations of oxygen (cerebral effects of anoxia), the

slightly to greatly increased irritability, the euphoric tendency to

laugh uproariously at meaningless trivia, the dull-witted phases, the

compulsive behavior at times and the comatose condition as the anoxia

approached the lethal stage and similar phases of some of the symptom

complexes studied in the hospitals (Knisely, Bloch, Eliot and 71arner,

1950, p. 104).

There is certainly justification for taking a long look at the

nervous system, particularly the cerebral cortex, as possibly one of the

most sludge-sensitive organs in the entire body. Perhaps all of the

blood which perfuses the brain must pass through the capillaries, for

there are apparently no arteriovenous shunts in the brain (Forbes, 195L).


XWhile in the phylogenetically older parts of the brain some anastomoses

are found (especially in the reticular formation) and may even form

extensive networks in pathologic conditions, these represent almost

exclusively interarterial connections (Klosovskii, 1963).

These anatomical considerations argue for an intimate relation-

ship between blood sludge and stagnant anoxia. If blood flow through

the brain is dependent upon the functional integrity of the capillary,

and if severe blood sludge plugs capillaries and induces a condition of

stagnant anoxia as Knisely (1951) has insisted, then blood sludge and

stagnant anoxia should be equated and the effects of blood sludge should

be the effects of stagnant anoxia.

Equating sludge with stagnant anoxia should permit some confident

predictions concerning the effects of sludge. It happens, however, that

this analogy is less restrictive than expected, for stagnant anoxia may

be equated with all other anoxias and therefore all that is known of an-

oxia may be relevant to the question of the effect of sludge on behavior.

The equivalence of all anoxias was pointed out long ago by Peters

and Van Slyke (1931). These investigators studied the effects of oxygen

deprivation under a variety of conditions and they reviewed the litera-

ture. They expressed their conclusions as follows: "One might expect

these different types of anoxias to retard oxidation predominantly in

different sets of tissues and produce correspondingly different outstand-

ing symptoms. But the recorded effects of general anoxias of varying

origins, including even the histotonic, impress one rather with their

similarity when the causative conditions are comparable severity, ra-

pidity of onset, and duration." (Peters and Van Slyke, 1931, p. 585).

With regard to stagnant anoxia specifically, Van Slyke pointed to the


fact that the behavioral symptoms described for cardiac failure where

stagnant anoxia may occur with a minimum of confounding "a sense of

exhaustion, dizziness, stupor, syncope, dull headache, memory impaired

especially for recent events, hallucinations, depression or exaltation,

disturbances of the special senses" (Levy, 1920, cited by Peters and

Van Slyke, 1931, p. 585) are essentially the same as those described

for anoxia due to altitude (Barcroft, 1920).

Anoxia and behavior. The brain has an extremely high

oxygen requirement. With perhaps 2 per cent of the body's total

weight it accounts for 20 per cent of the total oxygen uptake when the

body is at rest (Kety, 1955). The brain seems unable to tolerate any

measureable decrease in its normal level of oxygen uptake without the

development of mental symptoms of cerebral hypoxia. This relationship

between brain function and oxygen was summarized by Lassen (1959) who

concluded that "In all conditions of semicoma or coma which have been

studied whether due to anesthetics, acute hypoglycemia, apoplexy or

any other cause the reduction in consciousness correlates roughly

with the decrease of cerebral oxygen uptake regardless of the cause of

the acute cerebral disorder."

A variety of neurological and behavioral consequences have been

shown to occur with anoxia. In a study of the effects of acute anoxia,

Murder (1952) exposed rats to varying degrees of hypoxic anoxia and later

made cell counts of areas 10, 17 and 2h of the cerebral cortex. He found

reductions in cell count in all areas studied, and the relationship be-

tween level of anoxia and cell count varied with the location, remaining

constant for all levels in area 10, but increasing in areas 17 and 2h


with increasing exposure time. Further, he found that maze performance

error scores varied directly with cell counts in areas 17 and 2h.

Studies on the neurological effects of asphyxia indicate a dif-

ferential effect union excitatory and inhibitory systems. In 1939, Van

Harreveld and '.larmont found that cats whose spinal cords had been

asphyxiated for various periods of time showed exaggerated extensor tone

(after recovery) which usually lasted for the about three weeks preceding

death. It was suggested that the observed effect was due to selective

damage to inhibitory mechanisms. Additional support for this hypothesis

comes from observations that asphyxiation abolishes reciprocal inner-

vation (Van Harreveld, 1939).

Consistent with this hypothesis is the finding that hypoxia en-

hances the distinction between simple and choice reaction times. A

number of studies (Bauer, 1928) have demonstrated only a slight prolonga-

tion of simple reaction time from anoxic hypoxia at levels under 20,000

feet altitude which represents the approximate point of collapse for the

unadapted subject. On the other hand, choice reaction times were found

to be lengthened at a simulated altitude of about 16,000 feet (Jongbloed,

1935), and both accuracy and speed may be significantly affected at sim-

ulated altitudes above 15,000 feet (Tanaka, 1928) and 18,000 feet

(McFarland, 1932).

Sensory thresholds also appear to be responsive to anoxia. Stokes,

Chapman and Snith (19h8) investigated the effect of anoxia, hypoxia and

hypercapnia on cutaneous pain thresholds in man. There was no signifi-

cant effect from a 10 per cent oxygen mixture, but a 13 and 28 per cent

rise in threshold was obtained from 5 and 7.5 per cent mixtures, respec-

tively. In the rat a 7.5 per cent oxygen mixture reduced response to


thermal pain and a 5 per cent mixture eliminated it completely (Bullard

and Synder, 1961). A decrement in auditory sensitivity at low frequen-

cies and a slight enhancement at higher frequencies has been found with

reduced oxygen intake (Klein, Mendelson and Gallagher, 1961).

The extreme dependence on oxygen is consistent with the finding

that the brain meets its extremely high energy needs primarily through

aerobic glucose metabolism in both normal and pathological states (Las-

sen, 1959). A virtually continuous supply of oxygen is thought to be

critical for normal brain function. For example, the human brain con-

tains at any one time a total of only about 7 ml of oxygen which at the

normal resting rate of consumption of about 50 ml per minute would last

less than 10 seconds; and symptoms of anoxia would occur sometime before

complete exhaustion (Tower, 1958).

While the entire brain may show an increase in oxygen uptake dur-

ing a condition such as severe anxiety (King, Sokoloff and Wechsler,

1952), more commonly any change in metabolic activity remains a local

phenomenon because of the nature of the causal chain tying blood flow to

neural activity level. Localized alterations in blood flow occur with

activation of specific sites in the brain (Sokoloff, 1957). Increased

activity in the brain means increased energy requirements, and these

are met primarily through aerobic metabolism (Lassen, 1959). This in-

creased aerobic metabolism means increased production of CO2 and an

increased pCO2 of the venous blood, and cerebral vasodilation appears to

be almost completely dependent upon the pCO2 (Wyke, 1963).

However, under acute anoxia the brain will turn to anaerobic

metabolism with large increases in glucose utilization and CO2 produc-

tion (Tower, 1958). Although the efficiency of energy production


without oxygen is much less than when oxygen is provided, this capacity

for anaerobic metabolism should tend to give oxygen and glucose a consid-

erable measure of functional equivalence. Further, since pCO2 of the

venous blood determines the degree of vasodilation and since vasodila-

tion should govern flow rate, then all of these factors must necessarily

be involved in the organism's reaction to anoxia.

Brain oxygen and glucose relations. Different areas of

the brain vary in their sensitivity to the effects of oxygen or glucose

lack, but the changes in brain and behavior associated with deficiencies

of either of these two factors tend to be the same regardless of which

variable is manipulated. For example, Tschirgi (1960) has pointed out

that during acute anoxia the electrical activity of the cerebral cortex

survives for only It to 15 seconds, that of the caudate nucleus, 25 to

27 seconds, the ventromedial thalamus, 28 to 33 seconds and that of the

R. F. of the medulla from 30 to 0L seconds. Tschirgi's listing of the

course of brain anoxia follows closely the relationships defined by the

patterns of regional sensitivity to insulin shock. First there is a

depression of the cerebral hemispheres and of the cerebellum; then re-

lease of the subcortical diencephalon, the subcortical motor nuclei, the

thalamus and the hypothalamus, release of the mid brain, next the upper,

and the lower medulla (Himwich, 1952).

The intimacy of the relationship among CO2, 02 and glucose is also

reflected in the fact that a similar phasic response of the nervous sy-

stem occurs regardless of which variable is manipulated. There is a

three-stage reaction to CO2. With 3.5 to 7 per cent CO2 in respired air

there is some depression of cortical activity. From 5 to 20 per cent


produces a generalized reticular activation which reverses the original

depression. Finally, at levels above 25 per cent there is a general CO2

narcosis which is associated with the inactivation of the RAS (Wyke,


Some differences become evident, however, when other variables

are considered. Small, Weitzner and Nahas (1960) ventilated dogs with

5, 10 and 15 per cent CO2 and obtained large increases in cerebrospinal

fluid pressure (CSFP) within a '10 min. interval which lasted for as

long as 90 minutes without any consistent changes in arterial or venous

pressure at any time, until termination of the hypercapnia at which time

CSFP rapidly returned to normal. Ventilation with 8 per cent 02 brought

a more modest rise (84 per cent) in CSFP which approached its maximum

within 5 minutes accompanied by dramatic increases in arterial and ve-

nous pressures. Yet during 90 seconds of asphyxia there was a 175 per

cent rise in CSFP together with the same large increases in arterial

and venous pressures that were observed with the hypoxia alone. Sludg-

ing, if it reduces flow rate and plugs small vessels, should produce an

effect comparable at least to the combined effect of anoxia and hyper-

capnia as in asphyxia and not that of anoxia alone.

Other experimenters have been impressed by the similarity of func-

tional effects upon the brain of hypoxia and of hypoglycemia. Sugar and

Gerard (1938) pointed out that hypoglycemia may make an important con-

tribution to the damage associated with sudden anemias, while Gellhorn,

Ingraham and Moldavsky (1936) saw hypoxia and hypoglycemia acting syner-

gistically to produce convulsive seizures.

Wyke (1963) has suggested that the observed sedative effect of

hyperventilation associated with cerebral hypoxia is the result of

severe vasoconstriction of the brain. He also noted that these behavior-

al effects are correlated with slowing of the EEG and that both of these

effects are maximal during coincident hypoglycemia and hypoxia and mini-

mal when arterial concentrations of both 02 and glucose are high. McFar-

land and Forbes (1910) have found that the ingestion of glucose counter-

acted the tendency for the stimulus threshold for light to rise in

hypoxic anoxia. Perhaps the most dramatic demonstration of the func-

tional equivalence of oxygen and glucose and of the importance of blood

flow has been offered by Neely and Youmans (1963). They reported that

dogs survived 30 minutes of continuous perfusion of a buffered glucose

solution followed by replacement of the blood.

Brain blood flow and behavior. Blood flow rate appears

to be a critical variable in every consideration of brain oxygen-glucose

relations, for the transport of both oxygen and glucose and the removal

of 002 and of the acid residues of oxygen-free metabolism depend upon

the integrity of vascular function.

The involvement of blood flow changes in the response to cerebral

anoxia has been recognized and studied. For example, Geiger (1958) has

insisted that the nervous system's ability to tolerate a deficiency of

oxygen may depend upon the maintenance of a blood flow sufficient to

remove the toxic waste products produced by the anaerobic metabolism.

Gellhorn and Kessler (1942) have provided some experimental support for

the importance of blood flow in brain function. They made rats comatose

with injections of insulin following bilateral extirpation of the adre-

nal medulla. They applied electric shock to the brain and found that

this resulted at once in a recovery from coma, with a return of normal


behavior and normal EEG in spite of the fact that the blood glucose had

not risen above the coma level. It was hypothesized that stimulation

of the sympathetic NS resulted in an increase in blood flow to the brain.

Swank and Nakamura (1960) have reported that hamsters given sto-

mach loads of butter-fat become increasingly inactive during the hours

following the ingestion but become active again and develop convulsions

beginning approximately 3 hours after feeding. These behavioral changes

were correlated with decreases in the p02 of the brain tissue and with

increases in the viscosity of the blood; the p02 fell and blood visco-

sity increased. Swank and Escobar (1957) injected high molecular weight

dextran into dogs and obtained major increases in the viscosity of the

blood together with paralysis and electrocardiographic changes lasting

between 1 and'6 days. The most severe effects were obtained in those

animals which were given a long-acting anesthetic prior to the treat-

ment. It should be noted that the high molecular weight dextran also

produced sludging of the blood (Thorsen and Hint, 1950). Cullen and

Swank (195h) injected high molecular weight dextran into six hamsters

and observed extravascularization of trypan blue from both cortical and

subcortical vessels '6 hours later.

The consequences of acute cerebral circulatory arrest have been

reported by a number of investigators. Weinberger, Gibbon and Gibbon

(19h0) found permanent damage to the cerebral cortex in cats following

3 min. 10 sec. of complete circulatory arrest, while 3 min. 25 sec. pro-

duced a softening of the cortex. This sensitivity of the cerebral cor-

tex to impairment of blood flow to the brain has been a common finding

in acute experimental ischemia ('Wright, 1965). Dogs exposed to 6 min.

of apparently complete cerebral circulatory arrest have been found to


show no measurable behavioral decrement as determined by measures of

learning, retention or psychomotor performance, whereas 8 min. of circu-

latory arrest resulted in impaired learning but had no effect upon re-

tention or psychomotor performance (Nielson, Zimmerman and Cooliver,


Wright (1965) produced complete cessation of cerebral blood flow

in rabbits and cats. Only one of four cats survived 7.5 rain. and none

out of four survived 10 min. of cerebral vascular occlusion. On the

other hand none of three rabbits survived even 6 minutes. This dif-

ference might have been due to species differences in susceptibility to

the treatment. Wright suggests that the difference in survival expect-

ancy between the two groups of animals may be due to the difference in

the two methods used. Only the arterial inflow was occluded in the cats.

This treatment left the brain pale and bloodless. Both arterial and

venous flow was occluded in the rabbits, leaving the vascular beds of

the brain filled with blood for the entire period. Since more oxygen

and glucose should have been available to the brain in the latter in-

stance, the authors speculated that the lower survival time for the rab-

bits may have been due to thrombi forming in the small vessels of the

brain during the period of vascular occlusion.

The possibility that species differences in response to cerebral

vascular occlusion may have overshadowed any effects due to differences

in the techniques used is suggested by the finding as noted above that

dogs may tolerate 8 min. of cerebral circulatory arrest without severe

neurological damage while cats and rabbits could not even be revived

after 6 to 8 minutes. With regard to this question Wright (1965) has

pointed out that arterial network interconnecting with the carotid


system is extremely complex and that major differences in these systems

are comrnon both among species and among individual animals within any

given specie. Consequently the effect of the treatment may vary consid-

erably among individual animals while different species may actually

require different techniques for producing even a near total cerebral

ischemia. Typical maximum periods of reversible ischemia reported for

various species include 8 min. for the dog (Boyd and Connolly, 1961)

and 5 to 7 1/2 min. for the cat (Wiright and Ames, 196h), but no more

than 5 min. for the rabbit (Hirsch, Bolte, Schandig and Tonnis, 1957).

Cerebral blood flow conditions approximating those to be expected

during severe red cell aggregation may best be produced experimentally

by means of an extended period of less than total flow impairment. For

example, extensive infarction of the basal ganglia and cerebral hemi-

spheres together with gross behavioral deficits has been produced in the

dog by occluding the middle cerebral artery for a 2 hour period (Cyrus,

Close, Foster, Brown and Ellison, 1962). In those dogs which received

infusions of low molecular weight dextran before the occlusion the ex-

tent of tissue damage was significantly reduced and the behavioral

effects much less extreme. Since the effect of the dextran appears to

be that of reducing or preventing cellular aggregation and stasis of the

blood (Gelin, 1956), the differences in extent of tissue damage between

the controls and the dextran-treated animals should represent the ef-

fects of sludging and vascular stasis.

The above findings would seem to emphasize the role of thrombus

formation in the process of cerebral tissue damage during ischemia as

opposed to anoxia per se. Recent work by Neely and Youmans (1963) has

provided additional support for such a position. These experimenters


cleared the dog brain of blood by raising the CSFP to LOO mm Hg which

was at least twice that of the highest expected blood pressure level

and found that all of five animals exposed to up to 25 min. of such

treatment "were able to see, stand and hear the next day and survived

at least 18 hours." Six other animals whose brains were kept bloodless

for from 30 to 60 min. all had return of normal blood pressure and p02

of the arterial blood, but neurological abnormalities were evident and

none of the animals survived over 21 hours. Completeness of the ischem-

ia was verified by injecting radioactive sodium into the blood stream

and testing the brain for radioactivity 15 min. later. No activity was

present in the brain tissue as compared to a high activity for muscle

tissue overlying the brain.

The authors suggested that tissue damage associated with the

arrest of blood flow is related to the presence of anoxic blood in the

vessels and that the method used here prevented such damage through

clearing the blood from the brain before thrombi had time to form. Thus

the absence of blood served to protect the brain from injury. With glu-

cose available in the static blood the brain is presumed to produce

lactic acid via anaerobic metabolism of the glucose. The pH of the blood

is lowered and the clotting tendency is increased, an assumption con-

sistent with the finding by Crowell and Houston (1961) that even the

presence of heparin will not prevent the clotting of acedotic blood.

When no blood is present, this process is restricted, since brain tis-

sue lacks the stores of glycogen found in other tissues.

The results from the above study indicate that the brain is high-

ly tolerant of all manipulations except stasis of the blood. Yet the

permanent plugging of small vessels is precisely the function that


Anisely (i'^l) has ascribed to blood sludge. Tne brain then should tb

particularly sensitive to the effects of sl -.

'" There are re,'; reports of

v:ork deai1in; directly th the prob'lo. of a. itatin to slu ge. A ni.:-

ber of observati-s o: adaptive c:. s followi., c. oire to a.oxia

have boon reported. . '., otihesis of the functional equcivaloece of

sl-' -n arn anoxia :-.: .sts that najor a-'.-ive ;....,-es do occur durin

slud '-. hypothesis further su :. ts that the literature dealing

with adaptati.4i to anoxia should be relevant to the problem of adapta-

tion to a condition of blood sl... -.

increased resistance to h ia

may he produced :" a varied of conditions. ertlett an. P lps (1 .0)

showed thnt rats previously adapnte to sical restraint survived acute

.:'a better than controls, a relationship which held whehher th c ai

mals were restrain d or no, during the actual testir j. Acli natiti,

produced by discontinuous hy

cantly tne resistance to acute hypox

and I~ rr,ns, 19L6). Similar c...

(Lufu, 1 '; Van Liere and Stickn-y,

Prol .. exp, os're to ild 1

tolerance for acute hy oxia. r e

chronic res iratory i" nt "

inP arterial "cania an a oxia

rdict ecd for a noral sub .t '

L nSSn, 1 9).

has bean fo.d to increase si6i.if -

ia in rats ( Thor;n, C iton, barber

have been observed in :hu.as

1 r :.

)xia also results in a hen

C -pl>3, in hurr.as su, '-e11 fr"

.. r... n ' I.. 01T-

nsL4 i'r Iees oCf 'l G ci .1 --

si- 1.r levels of blocl ,sses

'o dobt a number of factors c -tribte. to the ada: tiv .


resulting from hypoxia. Van Liere and Stickney (1963) have suggested

that quantitative changes in metabolic activity of the tissues may rep-

resent one means of promoting adaption. This view finds support in a

recent report by Hamberger and Hyden (1963) to the effect that one of

the enzymes of the electron transport system cytochromee oxidase) showed

an increase of several hundred per cent following exposure for an extend-

ed period of time to a moderately reduced oxygen concentration in the

respired air.

Balke (194h) reported that in humans two weeks of training in a

mildly hypoxic environment of about 10,000 ft. altitude resulted in a

nearly one-third greater increase in maximum level of oxygen consumption

and capacity for work than could be obtained by training at sea level.

Failing to demonstrate changes in hemoglobin, myoglobin or pulmonary

ventilation which might have accounted for this increase, Balke conclud-

ed that the capillary blood flow in the muscles must have improved.

Since the testing situation was prolonged, the capacity of the subjects

to accumulate an oxygen debt such as has been found to concur with exer-

cise (Grolm-nan, 1955) should have had little or no influence on the


Adaptive changes in the blood-vascular system. A vari-

ety of evidence recommends that at least part of the mechanism of adap-

tation to anoxia involves changes in the size, shape and number of blood

vessels, especially capillaries. Such changes appear to be common dur-

ing the chronic and severe sludging which occurs in chronic diabetes

mellitus (Ditzel, 1955). Blood vessel changes associated with this con-

dition include capillary elongation tortuousnesss, twisting) and shifts


in arteriovenous ratios. These changes were found to compare closely

with those obtained in normals with the inhalation of 5 per cent CO2

mixtures (Ditzel, 196h).

The dependence of vascular changes in chronic sludging upon tis-

sue oxygen and CO2 levels is also suggested by the nature of the changes

occurring during adaptation to the relatively hypoxic conditions at high

altitudes. Liebesny (1922) reported that persons who travel from low to

high altitudes show changes in the condition of the blood as a result of

the altitude change. The blood flow is impeded, and the vessels, because

of the apparent clumping of the cells within, have a beaded appearance.

Simulated altitude studies have shown that hypertrophy (increases in

diameter and tortuosity) of the vasculature is the usual response to

hypoxia. In acclimated animals the capillaries may double in diameter.

The number of functional capillaries may also be increased such that the

total increase in vascularization becomes a function of both hypertrophy

and hyperplasia (Van Liere and Stickney, 1963; Korner, 1959). The net

effect of these changes in vascularity should be to increase the diffu-

sion gradients at the tissues and to permit continued integrity of func-

tion, particularly of the brain, at reduced levels of oxygen and other

nutrients and at the same time remove more rapidly any toxic waste


Also, Swank (1956) found that dogs receiving repeated fat meals

developed a "surprising" degree of adaptation which lasted for as long

as a year. As a result, instead of showing the usual increased blood

viscosity and red cell aggregation (Swank, 1951) following the ingestion

of the high fat meal, the animals showed essentially normal flow condi-

tions even after large test meals, and "remained active and vigorous".

(Swank, 1956).

Endocrine relations in adaptation to sludge. Endocrine

response during adaptative changes occurring as part of the stress syn-

drome may have a considerable effect upon the entire vascular system.

First, the proportions of the various molecular weights of plasma pro-

teins are sensitive to the circulating levels of adrenocortical steriods

(Ditzel, 1959), while the proportions of these protein fractions, the

relatively low molecular weight albumin and the higher molecular weight

globulins and fibrinogen, tend to determine the sludge level of the

blood (Ditzel, 1959; Fahraeus, 1929; Thorsen and Hint, 1950).

The adrenocortical hormones apparently affect the functional capa-

city of the blood vessels as well as the blood itself. The significance

of adrenocortical activity in producing and maintaining the functional

integrity of the blood-vascular system in the face of environmental

stressors is suggested by Zweifach's (1961) description of vascular

function in the adrenalectomized animal.

. a rise as small as 20 C. in the temperature of the fluid
bathing a tissue exposed for microscopy causes a rapid stasis
in the small venules and veins. Colloidal carbon in the circu-
lation accumulates along the intercellular borders of the capil-
laries, emphasizing the abnormal status of this constituent.
The capillaries undergo stasis following even minor mechanical
manipulation with micro-needles. The endothelium cells swell
and appear to have lost their normal tone. Numerous petechae
develop along the venous capillaries. Intravascular thrombi
fail to regress following even minor micro injury. Although top-
ical application of cortisone or of adrenal cortical extract
reverses within several hours the vasodilation and depressed vas-
cular reactivity typical of such vascular beds, the fragility of
the endothelium and the vessel wall as a whole is not signifi-
cantly lessened . (p. 109).

The effect of these hormones on blood flow has been demonstrated

in very dramatic fashion by Bergen, Hunt and Hoagland (1952) who showed

that a 61 per cent decrease in cerebral blood flow together with a

h6 per cent fall in brain oxygen consumption which developed in


adrenalectomized rats returned to normal in 2 to 3 hours following treat-

ment with lipoadrenal extract.

Adaptation. Comments. The restrictions upon the adap-

tive process should be repeated. Wyke (1963) has pointed out that with

continued exposure to high levels of CO2 in the respired air there is a

gradual reduction of symptoms, indicating adaption to the highly abnor-

mal conditions. In anoxia, however, the symptoms may change with time,

but they do not go away. The effects of acute anoxia are suggestive of

alcohol intoxication, whereas chronic anoxia produces a condition resem-

bling physical and mental fatigue (Barcroft, 1920). The equating of the

effects of acute anoxia with alcohol intoxication seems rather appro-

priate here, since alcohol is reported to be an effective agent for the

production of intravascular aggregation (Bloch, 1956; Knisely, 1951).

This latter observation suggests that any adaptive advantages gained

from anoxia may also accrue to the periodic user of alcohol.

In summary, the evidence available with regard to adaptation

seems to offer a strong argument for the dependence of sludge effects

upon the prior stress history of the animal. Control for this factor

seems to be recommended for any study concerned with the effects of


Present Status of the Sludge Problem

One observation which may be derived from the preceding pages is

that there is a dearth of experimental data bearing directly on the

question of the significance of sludge. This deficiency represents the

basic criticism of the. present status of the sludge problem.

The criticism is hardly original. Laufman (1951) called for more


data when he insisted that it should be first determined whether sludge

has any detrimental effects on the organism before the question of how

is taken seriously. While Laufman is to be commended for his concern

with objectivity, his failure to see these two questions as representing

two levels of analysis of the same problem instead of two separate prob-

lems may have added to the confusion. Ditzel (1959) also called for

more data when he pointed out that "our knowledge concerning the patho-

physiological significance of pronounced intravascular aggregation per

se is inadequate, and concentrated effort should be exerted for the

elucidation of this important problem," (p. 57).

There was only a very limited response to this latter appeal.

The difficulty lay in the fact that the question was not "whether" to

attack the sludge problem, but "how" to attack it. This difficulty has

recently been summarized by David Long (1962).

The pathophysiologic relevance of intravascular aggregation has
been disputed for years. This dispute will probably continue for
as long as our methods for evaluating function are so gross and
imprecise for organs with a large functional reserve. Further-
more there are no methods for quantifying the magnitude of intra-
vascular aggregation of blood corpuscles, In the meantime micro-
circulationists remain impressed when large aggregates of cor-
puscles occlude arterioles and venues preventing capillary
circulation to focal areas. (p. 579).

Since behavioral studies on the problem of blood sludge are almost

nonexistent, the present status of the sludge problem should represent

even more of a challenge for the psychologist than for the "microcir-



An extensive literature search produced no reports of systematic

observations made on behavior during experimentally controlled blood

sludge. In the absence of such studies, an effort was made to formulate

some predictions about the behavioral consequences of sludge through

defining the mechanism of the action of sludge.

Evidence was offered for the following argument. Blood sludge

promotes a condition of stagnant anoxia through impaired blood flow and

the plugging of small vessels. The behavioral effects of all anoxias,

including stagnant anoxia, are the same, but the effects vary as a func-

tion of intensity, rate of onset and duration. In particular, major

changes in response to anoxia seem to occur with adaption during pro-

longed exposure. However, so long as anoxia is involved in severe blood

sludge, that condition should be reflected in behavior.

Accordingly it was predicted that severe sludge in the rabbit is

associated with behavioral changes which could range from a moderate de-

pression of activity with increased response latencies and short term

memory deficits to a more severe condition approaching stupor or even


Attempts were made to test this hypothesis as well as to make

observations on a variety of other behaviors which might be sensitive to

the effects of sludge. These efforts are described below in three parts.

Part I. The problems associated with the development of a method


for the study of sludge-behavior relations are discussed together with

the solutions to these problems.

Part II. Results from two separate experiments are offered as

evidence for the validity and reliability of an in vitro technique for

the evaluation of the in vivo condition of the blood.

Part III. The relationship between blood sludge and behavior is

explored through a series of four experiments covering a variety of be-

haviors beginning with simple maintenance behaviors and extending to

behaviors of increasing complexity.

Part I. Preliminary Problems

The four requisites for the study of the behavioral correlates of

red cell aggregation are 1) an appropriate subject, 2) a means for induc-

ing the aggregation, 3) a method for measuring the degree of aggregation

and, finally, h) a behavioral test.

These four factors, together with the attempts to specify and com-

bine them, are discussed below.

Appropriate subject

The rat might have been the preferred subject for the present

series of experiments because of the large number of standardized behav-

ioral testing situations already available for use with this animal.

However, the rat has what appears to be a species specific sensitivity

to dextran (Adamkiewicz and Adamkiewicz, 1960), since only a small per-

centage of animals from a few colonies of Wistar rats have been found to

be non-reactors (Harris and "West, 1963). Further, use of highly viscous

solutions with the rat requires anesthesia and surgical manipulation.


Because the anesthesia and the high molecular weight dextran may have

complementary effects,the use of the rat in studying the effects of

sludge on behavior might make increasingly difficult the task of isola-

ting the effects of the sludge per se.

A non-rodent, the rabbit, (order Lagomorpha),was finally chosen

as the best possible compromise. The ear veins of the rabbit are acces--

sible without the need for anesthetics or surgical manipulation and they

are large enough even in an animal of 750 gms body weight to tolerate

continuous infusions of the highly viscous dextran solutions. The number

of alternative sites for access to the vascular system available in the

ears of the rabbit make the animal a particularly attractive choice.

Successive infusions or blood samples can be scheduled with little risk

of exhausting all alternative infusion sites before the experiment is

completed. Further, infusion and blood sampling techniques have been

developed to a point where the animal seems to be almost completely un-

mindful of the treatments and as a result there is minimal confounding

from incidental aspects of experimental manipulations.

The choice of the rabbit was also recommended by the fact that

there is already a considerable amount of information available on the

use of this animal in the study of red cell aggregation and particularly

in experiments where the aggregation is induced by infusions of high

molecular weight dextran (Fajers ard Gelin, 1959; Gelin, 1956; Gelin,

1959; Gelin, 1962; Lofstrom, 1959; Lofstromand Zederfeldt, 1957; Thorsen

and Hint, 1950; Zederfeldt, 1957). In none of these studies has any

unusual sensitivity to the dextran been reported.

A method for inducing the sludging

As a means for inducing the aggregation high molecular weight

dextran (HmDx) seems to have obvious advantages. First, except when

used with the rat, there appears to be a minimum of confounding second-

ary effects. As previously noted, no hypersensitivity had been reported

in any of the studies involving the use of dextran with the rabbit.

Infusion of HmDx (2,000k mean molecular weight) into rabbits has

been shown to induce a number of physiological effects. The most drama-

tic effect of the infusion of HmDx into the blood stream appears to be

the aggregation of the red blood cells, the severity of which is deter-

mined primarily by two easily controlled factors, dose level and mole-

cular weight of the dextran (Gelin, 1956; Thorsen and Hint, 1950).

HmDx appears to have a prolonged effect upon the condition of the

blood. Gelin (1959) has reported that a single injection of 1 gm/kg

body weight in the rabbit induced severe sludge, increased blood visco-

sity and heightened ESR which lasted for $ to 6 days. The duration of

the effect of HmDx is presumed to be related to the fact that it is re-

moved only slowly from the vascular system.

There appears to be little loss into the extravascular compart-

ment of colloidal materials above a molecular weight of hl2,000 (Mayer-

son, Wolfram, Shirley and Wasserman, 1960). Normally the dextran not

lost through the kidney seems to be taken up gradually by various cells

in the body and eventually broken down into glucose. Most of this dex-

tran appears to be removed from the circulation by the reticuloendothe-

lial system of the liver and spleen (Grotte, 1956; Osol and Farrar, 1955).

However, when these phagocytic reticuloendothelial cells become saturated


following large doses of foreign colloidal particulate matter (such as

dextran) the endothelial cells in the capillaries and venules in other

areas of the body may become phagocytic. Because these endothelial

cells are unable to metabolize the ingested particles, the results of

the infusion of foreign colloidal material into the blood stream may be

a generalized pathological condition of the entire vascular system

(Zweifach, 1961). Support for this view is offered by the finding that

a single injection of HmDx into the rabbit produces no obvious patholo-

gical lesions of the viscera even though the sludging remains severe for

several days whereas successive injections of HaDx have consistently

produced tissue damage to liver, kidneys, heart and lungs (Gelin, 1956;

Gelin, 1959) or seriously impaired the healing to wounds (Zederfeldt,


There is some evidence, however, that the phagocytic response may

interact with the sludging effect. Knisely, Bloch and garner (1948) have

reported that when particles of India ink are injected into the vascular

system of the frog they become covered with a "sticky coating" and are

thereupon ingested immediately upon contact with the phagocytes of the

liver sinosoids. A similar fate apparently befalls the masses of para-

sitic red cells of the monkey during Stage III of Knowlesi malaria.

These cell masses are also phagocytized in the liver. Knisely et al.

(1948) estimated that in one case the rate of destruction was sufficient

to have destroyed within a 3 hour period up to one third of all circula-

ting red cells.

On the other hand, it seems that the mere presence of the foreign

colloids in the blood stream does not result in pathologic changes in

the vascular endothelium, or at least none severe enough to promote


damage to surrounding tissues. For example, Fajers and Gelin (1959)

found that additional dextran infusions in the form of solutions of low

molecular weight dextran (L-nDx) prevented the tissue damage produced by

the infusions of HmDx alone, and Zederfeldt (1957) found no interference

with wound healing when LmDx infusions were given together with dose

levels of {HDx which when given alone over four successive days caused

significant retardation of healing.

Gelin (1956) has reported, also, that anemia occurs with severe

red cell aggregation whether the aggregation is induced by tissue trauma

or by infusions of HmDx, although he related the reduction in circulating

cell volume only to vascular stasis and the packing of cells in capilla-

ries and venules. There does seem to be some tendency toward specificity

in this reaction. For example, the introduction of bacterial endotoxins

into the vascular system is reported to increase phagocytic activity and

tissue damage in the endothelia of the lungs, kidneys, adrenals and in-

testines (Zweifach, 1961), those organs which Fajers and Gelin (1959)

found to suffer damage during intravascular aggregation induced by in-

fusions of IHmDx.

All dextrans have been observed to increase the apparent viscos-

ity of the blood. The magnitude of this increase appears to vary

directly with the molecular weight of the dextran used (Gregersen, Peric,

Usami, Chien,Chang and Sinclair, 1963). Further, the dependence of

blood viscosity upon shear rate also seems to vary directly with the

molecular weight of the dextran (Rand and Lacombe, 196h).

Some of the effect of the HmDx appears to be related to the fact

that it also acts as a plasma expander. For example, Gelin (1956) found

that following a single injection of 1 gm of HmDx the urinary output fell


from a preinjection level of 2h0 ml per day to a low of 20 ml on the

first posttreatment day and rose gradually thereafter to 75 ml on the

second day and to 150 ml on the third day after the injection.

When HmDx is used for inducing sludging, there are available two

methods for isolating those effects of the dextran which may be independ-

ent of the effect of the induced aggregation. First, a separate con-

trol group receiving infusions of a dextran with a molecular weight of

between 75,000 and 80,000 may be used. Second, the blood sludge induced

by the infusion of HImDx may be reversed by treatment with a low molecu-

lar weight dextran (LmDx) having a molecular weight of less than h0,000.

Dextran of a molecular weight between 70,000 and 80,000 (approx-

imately that of blood albumin) does not induce red cell aggregation.

However, it does act as a plasma expander, and so it may therefore be

used to control for the plasma expander effect of HmDx. In normal sub-

jects there is no significant effect of erythrocyte sedimentation rate

(ESR) and perhaps only a slight rise in blood flow associated with a

minor decrease in peripheral resistance while subjects with high sedi-

mentation rates show little if any reaction to infusion of dextran of

this weight other than perhaps an extremely slight increase in plasma

viscosity (Gelin anJ Thoren, 1961).

LTDx also acts as a plasma expander, but its effect is shorter

and more dramatic than for the higher molecular weight dextrans. The

effect lasts only about 1 1/2 hours for a 10 per cent solution and not

more than 3 hours with a 15 per cent solution when the dose volume is

held constant. As a result, a significant concentration of LmDx remains

in the blood stream for a much longer period, and so the flow-promoting,

sludge-reducing effects of LzDx should last for a somewhat longer period


(Gelin, Solvell and Zederfeldt, 1961).

Together with the plasma expansion there is a marked increase in

diuresis, but no significant urinary losses of sodium or chloride nor

changes in blood levels of these factors when salt-free solutions were

used (Gelin, Persson and Zederfeldt, 1961). Excretion of the LmDx in

the urine is rapid at first, but drops suddenly to extremely low levels

even before the diuresis is reduced. Finally, no significant changes

have been observed in the levels of the various plasma protein fractions

during this period (Gelin, Solvell and Zederfeldt, 1961).

The determination of the effect of the dextran per se should re-

quire also an evaluation of the influence of the large doses of saline

solution which serve as the vehicle for the dextran. The saline appears

to have no significant plasma expansion effect (Gelin, S6lvell and Zeder-

feldt, 1961). Neither have any measurable pathological effects been

found from frequent and successive infusions of saline solutions (Zeder-

feldt, 1957). Nor does it seem to have any effect upon HmDx-induced

red cell aggregation when the saline is used in place of the LmDx (Cullen

and Swank, 1954), or upon the sludging occurring during eclampsia which

has been reduced by intravenous albumin (Odell, Aragon and Pottinger,

1947). On the other hand, Brooks, Dragstedt, Warner and Knisely (1950)

have reported some apparent improvement from saline infusions in sludg-

ing following severe thermal burns. Robbins (1963-1961) found a reduc-

tion in survival expectancy in saline injected rats made to swim to

exhaustion. Further, preliminary observations with rabbits have indi-

cated a depression in spontaneous activity following saline infusions

in volumes required to match those of the dextran injections.

Since no conclusive evidence is available regarding the effect of


infusions of physiological saline into the rabbit, the present series of

studies provided for control of this factor in every instance. In every

experiment, the effects of dextran per se were also controlled by means

of the first of the two methods described, that of matching the volumes

of HmDx infusions with solutions of dextran with a molecular weight

approximating that of blood albumin (77 k molecular weight).

Evaluation of level of sludge

Measurement of the degree of aggregation of the red cells has

long been recognized as perhaps the most serious obstacle to the study

of the effects of the aggregation. Fahraeus (1929) equated intravascu-

lar aggregation with erythrocyte sedimentation rate (ESR) but a number

of more recent studies indicate that the two measures have only a limited

correspondence (Bigelow, Heimbecker and Harrison, 19h9; Ditzel, 1955;

Gelin, 1959).

Knisely (1951) has shown a continuing dependence on direct in

vivo observations of blood flow. He and his coworkers have emphasized

that the blood flowing through any single artery is a statistically valid

sample of all the blood in the body (Bloch, 1953; Bloch, 1956; Knisely,

Bloch, Eliot and Warner, 1950; Knisely, 1951). Exceptions to this con-

clusion are to be expected only in cases of severe sludging where consid-

erable settling of cellular elements may occur (Knisely, 1961). The

observations on which this conclusion is based were made on several spe-

cies and large numbers of animals, and the tissues studied included the

intestinal mucosa, the mesenteries, liver and the surface of the brain;

and it was concluded from these observations that "at all times the

blood coming down the arterioles of uninjured bulbar conjunctiva is a


statistically valid sample of all the flowing arteriole blood in the

body" (Knisely, Eliot and Bloch, 1945, p. 221).

It may be possible, however, that Knisely and his group never

made simultaneous observations of both mesentery and conjunctiva during

severe sludging. At least they do not state specifically that they had

done so (Knisely, Eliot and Bloch, 1945). On the other had, Bloch

(1956) reported finding no differences in the appearance of the blood

flow between these two tissues in the rabbit, cat, dog or rhesus mon-

key while under nembutal anesthesia nor did he find any changes in blood

flow in conjunctival vessels as a result of the anesthesia; but he made

no mention of the condition of the blood in these animals at the time of

the observation. It may be assumed that these were all "normal" animals

since Heimbecker and Bigelow (1950) found from the simultaneous compar-

ison of these two areas in the rabbit during severe induced sludging that

in the rabbit "the degree of sludging in the conjunctiva and nicititating

membrane was invariably more marked than that observed in the momentum

and mesentery". A number of such comparisons made by the present

writer on the rabbit and rat have tended to support Heimbecker and Big-

elow's (1950) findings.

Yet if the objective were that of validating in vitro measures by

in vivo observations then any differences in observed severity of sludg-

ing among the various tissues should be immaterial so long as the in-

duced changes in level of aggregation vary in a consistent manner in any

given tissue.

Actual determination of the condition of the blood via in vivo

observation in the experimental animal presents a dual problem. The

observation almost invariably represents a highly traumatic experience


for the animal, and it can be expected to affect the condition of the

blood flow as a consequence, thereby at the same time setting up highly

abnormal conditions for viewing the aggregation since changes in the

flow rate appear to influence the observable level of aggregation

(Ditzel, 1959; Hirschboek and Woo, 1950). Thus, any differences in

handling the animals might be reflected in differences in the apparent

level of in vivo aggregation. In addition, the actual observation is

made difficult by the rapid breathing and the struggling to escape. The

method is therefore seriously inadequate for use in the study of behav-

ioral variables.

The use of an anesthetic might seem to offer a partial solution

to the problem. However, there are indications that recovery from the

anesthetic varies with the severity of the induced aggregation (Robbins,

1963-196!; Swank and Escobar, 1957). Further, successive observations

would necessitate successive applications of anesthetic, with an in-

creased probability of confounding.

The greatest problem, as David Long (1962) has recently pointed

out, is met in trying to evaluate that which is finally observed. The

usual approach has been that of using either four or five separate cate-

gories based on some criteria, presumably size of aggregates, which left

a great deal to subjective judgment (Hirschboek and Woo, 1950; Odell,

Aragon and Pottinger, 1947; Robertson, Wolf and Wolff, 1950). As Hirsch-

boek and Woo (1950) have pointed out, the arbitrary standards used for

in vivo quantification of sludge are "quite crude", and while the method

may be of value when applied by a single observer it remains extremely

difficult to make any kind of meaningful comparison among observers.

Dissatisfaction with these highly subjective approaches promoted


development of a grading system for red cell aggregation which involved

four categories of aggregation and which also took into account the

location of the aggregation, that is, whether the clumping occurred in

the venules or arterioles (Ditzel, 1955; Ditzel, 1959; Ditzel and Sagild,


It is felt, however, that in spite of the attention paid to the

development of in vivo methods for evaluating sludge any such technique

must involve a high degree of subjectivity.

One alternative to direct observation of the blood flow seems to

be the use of blood smears. The presence of red cell aggregation has

been determined by this method for many years (Fahraeus, 1929; Hunter,

1835; Jones, 1843; Lister, 1858; Norris, 1869). In fact, Jones (1843)

even saw in the blood slide a means for estimating the sedimentation

rate, since he pointed out that "in order to know if a patient's blood

has a buffy tendency or not, it is sufficient to take a drop of blood

from the finger tip, press it between two pieces of glass and observe

whether or not it shows a dotted appearance. And just as there are dif-

ferences in the degree to which the buffy coat is developed, correspond-

ing degrees may be distinguished by the distinctiveness with which these

dots make their appearance." (Jones, 1843, cited by: Fahraeus, 1929,

p. 244). The "buffy layer" refers to the layer of fibrin which appears

on top of the column of red cells when blood with a high ESR is allowed

to stand so that the clumped cells settle to the bottom of the tube be-

fore clotting can occur (Fahraeus, 1929).

,"fhile there are no known reported instances of a complete depend-

ence upon this method for the quantitative evaluation of red cell aggre-

gation, a number of investigators have used the method to validate

L L;

observations made with other techniques (Fahraeus, 1929; Fahraeus, 1958;

Gelin, 1956; Thorsen and Hint, 1950). Gelin (1956, 1962) compared blood

flow in the conjunctiva of the rabbit with blood smears and found what

appeared to be a close relationship. The comparisons were made before

and after reversal of :.-,D::-induced sludging by means of infusion of LmDx.

Fahraeus (1958), too, was impressed with the possibilities of this tech-

nique for the evaluation of the condition of the blood. Extensive pre-

liminary work by the present author has indicated that the method may

offer a higher degree of reliability and objectivity and simplicity than

can be achieved by any other means. The validity of this assumption is

demonstrated in the first of the present series of experiments.

Behavioral tests

The decision to use the rabbit brought with it the problem of

defining suitable behavioral testing situations. Yhile the rabbit has

been highly recommended as a subject for behavioral studies (Zarrow,

Sawin, Ross and Denenberg, 1962), most of the work dealing with rabbits

has involved classical conditioning experiments (Fromer, 1963; Schnei-

derman and Gormezano, 1961) and observations of maternal behavior

(Deutsch, 1957; Sawin, Denenberg, Ross, Hafter and Zarrow, 1960).

The neglect of the rabbit in behavioral research may be related

largely to its apparent timidity. Fink (195)), who tested a single rab-

bit on his "Arrow" maze during a comparative study of maze performance

which ranged from turtle to man, found the rabbit to show an extreme

degree of initial timidity in the testing situation. However, in spite

of the animal's initial reaction to the testing situation, its total

performance in the maze placed it at a level between the rat and the cat.


More recently Livesey (196h) tested three rabbits (sophisticated) on a

double alternation problem in the WGTA and found their overall perform-

ance to be comparable to that of cats tested under similar conditions

(Stewart and Warren, 1957). Livesey (1965) also compared directly the

performance of rabbits, rats and cats on the double alternation problem

and again found the performance of the rabbit to compare favorably with

that of the cat, while the rat required approximately twice as many

trials to reach criterion of 60 per cent correct responses over 50 con-

secutive series of responses as did the rabbit or the cat. DeBaron

(1962) successfully used rabbits in an instrumental conditioning situa-

tion in which the animals were required to discriminate between a con-

tinuous and an intermittent buzzer. Voronin and Napalkov (1959) have

reported using the rabbit in conditioning response chains containing up

to seven or eight links, with only the final link reinforced by the US

(food). Jackson (1965) has conditioned chained responses in a number of

rabbits intended for exhibition purposes.

No other studies involving conditioned behavior in the rabbit

have been found. Consequently it was necessary to develop suitable test-

ing situations for this animal which is not only exceptionally timid but

extremely curious as well.

Part II. Measuring the Blood Sludge

An extensive series of preliminary observations had indicated

that the appearance of the blood smear obtained and viewed under stand-

ard conditions varies in a consistent manner with the condition of the

blood in vivo. Accordingly it was decided to verify this relationship

by means of systematic comparisons among in vivo and in vitro observations


made under standard conditions. The relationship is demonstrated through

a series of two experiments, each emphasizing different aspects of the


Experiment 1 was designed to test the validity of the blood smear

technique as a means for estimating the intravascular aggregation tenden-

cy of the blood. In vivo and in vitro photographic observations were

made before and after treatment with saline, 77k dextran or HmDx. Com-

parisons among matching sets of these photographs showed that the condi-

tion of the smear varies in a consistent fashion with changes in the in

vivo appearance of the blood. This experiment also provided an opportu-

nity to make direct comparisons on the in vivo condition of the blood

following each of the several treatments.

Experiment 2 was designed (a) to test the reproducibility of the

results obtained with the blood smear technique for the evaluation of

sludge, and (b) to provide standards for a blood smear rating scale.

Sets of two blood smears each were obtained following successive frac-

tional doses of sludge-inducing HmDx. Photographs obtained from two

locations on each slide from each set of slides were compared for a dem-

onstration of the reliability of the technique. Four photographs were

chosen as standards to represent blood sludge levels ranging from no

sludge (+1) to severe sludge (+4).

The two experiments are presented below.

Experiment 1. Validation of the blood smear technique by comparisons of
in vivo and in vitro observations


Subjects. Four albino rabbits, approximately 20 weeks old and


weighing between 2500 and 2800 gms, were tested by means of blood smears

for the absence of significant levels of blood sludge and assigned at

random to four treatment conditions, the same conditions used throughout

the entire series of experiments: normal control, saline-injected con-

trol, 77k dextran control and the HmDx-treated (sludged) condition.

Apparatus. The motion pictures of blood flow were taken on

Kodachrome II film at 64 frames per second using a 16 mm reflex camera

and a Leitz ophthalmic binocular bi-objective microscope with a 12X ob-

jective and an 18X eyepiece. The tissue being photographed was trans-

illuminated with a Knisely (1936) type quartz rod illuminator equipped

with a 1000 watt light source and a water heat filter. The end of the

5/8 in. diameter quartz rod was placed 2 in. from the exposed tissue.

Attached to the microscope stand was a paraffin lined cork board with a

2 cm. hole over which a section of the mesentery was pinned for obser-

vation. A small adjustable stand was used to hold the animal in posi-

tion under the cork board. All observations were made with the micro-

scope in the horizontal position, an observational technique highly

recommended by Knisely and his coworkers (Knisely, Warner and Harding,


Blood smears were photographed on Ansco Hypan film with a Leica

35 mm camera on a tri-ocular microscope using a 40 X dry objective, a

10X eyepiece and a Mikas micro attachment with a 1/3 intermediate

adapter. Areas of standard density were located on the blood smears

with the use of a photometer and densitometor equipped with a 1.96 mm


Animals were anesthetized with hexabarbital sodium ("Svipal


sodium") dissolved in tap water to a concentration of 50 mg per cc and

applied intravenously. The HmDx used for inducing the sludge was a

2,000,000 molecular weight dextran obtained from Pharmacia, Uppsala,

Sweden. The 77,000 molecular weight dextran was obtained from Pharma-

chem Corp., Bethlehem, Pennsylvania. Both dextrans were prepared in 50

cc lots as 10 per cent solutions in normal saline and all solutions

were autoclaved for h5 min. at 1180 C. before use. Each lot was used

for a single treatment only, and it was discarded and replaced if not

used within 5 days. Both the treatments and the anesthetic were

administered through a 23 gauge needle joined to a 'two-.way valve by a

50 cm length of 0.023 polyethylene ("Intramedic") tubing. A constant

temperature water bath was used to maintain all solutions at approximate-

ly the mean body temperature of the animal (h00 C.)

Procedure. Each animal was first anesthetized. A 10 cm area

was shaved along the ventral midline where a 6 cm incision was made in a

bloodless operation using surgical scissors. The animal was placed on

the adjustable stand under the cork board attached to the microscope

stand and a section of the mesentery suitable for 'observation and pho-

tography was pinned to the cork board with the desired field of view

extending across the hole in the board. The exposed tissue was bathed

continually with isothermic saline throughout the entire period of


The schedule of treatments and observations was as follows. Ten

feet of film was exposed as soon as camera and microscope were properly

focused. The appropriate treatment was then administered during a 10 to

12 min. interval. An additional 10 ft. of film was exposed one hour


after the beginning of the treatment. During this one hour period the

position of the microscope and camera remained unchanged, thereby per-

mitting a comparison of treatment effects against a constant background.

Blood smears were taken immediately after each film series and were coded

for later processing.

Areas of standard density were located and marked for each of the

smears. The blood slides were stained for 4h sec. using Wright's stain

followed by application of buffer. After the slides were dry, 35 mm

photographs were taken from the marked areas. A matching set of photo-

graphs representing in vivo observations was made by enlarging selected

single frames from each of the 10 ft. sections of exposed 16 mm film.


Acceptable motion pictures showing the condition of the blood

before and after treatment were obtained for the three animals receiving

infusions of the different solutions, but not from the normal control

animal. Although normal blood flow was maintained in the untreated

animal throughout the entire period as determined by direct microscopic

observation of the mesenteric vessels, camera malfunction rendered the

exposed film useless.

All blood smears taken from the three treated animals were suc-

cessfully processed and photographed. These photographs, matched with

enlargements obtained from the 16 mm film, are presented in Fig. 12

(Appendix) to permit (a) comparison of the in vivo condition of the blood

and the appearance of the blood smear, and (b) comparisons of the condi-

tion of the blood following each of the several treatments.

Comparisons among these photographs show the following. All


photographs taken during the pretreatment period show a complete absence

of blood clumping either in the vessels or on the smears. After treat-

ment with saline neither observation indicated any change from pretreat-

ment conditions. The posttreatment photograph for the 77k dextran

treatment shows a slight aggregation tendency for the blood smear. The

matching photograph (in vivo observation) for this condition shows in-

creased streaking of the particles present in the blood stream, indicat-

ing an increase in rate of flow. The distribution and the size of these

particles suggest no major change in intravascular blood sludge level

following treatment. By contrast, both photographs for the posttreatment

EHDx condition show a severe aggregation of the blood. The smear shows

large formless masses in which the outlines of the individual cells are

lost, while the in vivo observation reveals large cell masses separated

by spaces containing only plasma. Also, the cell masses show a tendency

to settle at the bottom of the vessel leaving only plasma at the top.


Comparison of the in vivo and the in vitro observations on the

condition of the blood before and after each of the three treatments

revealed a close correspondence between the two sets of observations

under all conditions studied. This close correspondence is regarded as

a demonstration of the validity of the blood smear technique as a means

for evaluating the in vivo aggregation tendency of the blood.

Both sets of photographs revealed major changes in the condition

of the blood following treatment with HmDx but not after either of the

other treatments. These changes consist of the presence of large cellu-

lar aggregates in the blood stream, dramatic slowing of blood flow and


settling of the cellular aggregates to the bottom side of the vessel.

This difference in the appearance of the blood which occurred

between the HmDx condition and both controls is regarded as a demon-

stration of the adequacy of the present technique for the control of the

independent variable, the level of blood sludge.

It should be pointed out that while only one animal was used for

each of the three conditions in the present study, the results from these

observations were found to be entirely consistent with those from a large

number of preliminary observations made under comparable conditions.

Experiment 2. Demonstration of the reliability of the blood smear tech-
nique and development of a rating scale for evaluation of blood smears


A single normal sludge-free unanesthetized male albino rabbit

weighing approximately 2300 gms was given the standard 1 gm/kg body

weight dose of HmDx in the form of ten equal injections at 12 min. inter-

vals, a period of time considered sufficient to assure an essentially

complete mixing of the blood following each injection (Pritchard, Moir

and MacIntyre, 1955). The infusions were made into the marginal vein of

one ear. Two blood smears were taken from the marginal vein of the other

ear immediately prior to each injection to permit comparison of the con-

dition of the blood smear with the magnitude of the cumulative HmDx dose.

Each slide was coded for later identification. The two sets of slides

were processed in the manner described in Experiment 1. Two points of

standard density were identified on each smear. A single photograph was

taken of each of the two locations on each smear from each set of slides.

From the eleven sets of photographs representing all dose levels, four


sets were chosen to serve as standards for use in rating the blood

smears taken from those animals serving as subjects in the several

studies on the behavi ral consequences of sludge.


Tn .~ tion of the eleven sets of togra ps taker fro: blood

smears obtained idring the course of the successive infusions of .x

revealed a hirh ree of si ilariZt a..,.: the four t, tra'phs of each

set but major differences .., the several sets.

The differences a- ,, the sets were found to vary consistently

with HmDx dose. Begi.viin, with (1) a random distribution of iA.ivciual

cells before treatment there is evident an increasing (2) clusterir. of

cells into rouleaux, followed by the development of (3) networks of

rouleaux masses interconnected Ly chains of rouleaux with the i'.. 1vi d

cells 7- :-' ,ore difficult to identify in the cell masses, unt'l fi-

nally with maximum dose levels the interco ..cting rouleaLx formcaions

become r, -' and abead or are .-ssin,; alt theirr an: the ..'anca of

the smear is that of (L) formless isolated rass:s within which the out-

lines of the individual cells an of the ro;lea;x are lost.

Photo raphs most nearly .- ... ive of the con ti7os descri:

above 'ere fIoun to be those obtain .d fr,: s-ears taken (I) before treat-

nent, (2) after 20 per cent of st9 nar o, (3) after -r cent o

standard dose nd (t) after 1 i a:r cent of tho s.ar"I dose .2 I i- of

H:Dx par bo:iy wei, .t. a four sobs of .,lst y

or dr : increas'-. severity of blood si a, are present in i. 13.

The other seven sets of phot showed tie sa .e internal consistency

found in the above four and thov were excluded onl to avoid overburd


this report with an excess of detail.


Each set of four photographs, representing two observations per

smear from each pair of smears, were found to show an almost complete

absence of variability while major differences occurred among successive

sets of photographs. These findings indicate that (1) the appearance of

the individual blood smear is highly uniform when thickness of smear is

held constant, that (2) random variability among successive blood smears

is very low when the aggregation tendency of the blood is held constant,

and that (3) the reproducibility of the results using blood smears is

independent of the blood sludge level at the time of the observation.

The four sets of photographs which were presented in Fig. 13 rep-

resent the four standard sludge levels for the blood smear rating scale.

These levels might be described as normal, mild, moderate and severe and

they are designated as sludge levels +1 through +-, respectively.

The use of only four categories of blood sludge maximizes inter-

class differences and should make the scale both simple to use and effec-

tive as a tool for the task which has been regarded as perhaps the great-

est problem in research on blood sludge the quantification of the

aggregation tendency of the blood.

Part III. Behavioral Studies

With standardized techniques available for the production and

measurement of blood sludge in the rabbit and with meaningful behavioral

testing situations defined and modified for use with the rabbit, the

exploration of the relationship between sludge and behavior was carried


out through a series of four separate experiments.

Experiment 1. Observations on the effects of HmDx-induced blood

sludge on food intake, water intake, spontaneous home-cage activity and

exploratory behavior.

Experiment 2. A study of the effects of blood sludge on response

latency in a shock-escape learning situation.

Experiment 3. A study of the effect of sludging on learning be-

havior defined in terms of response latency and running time in a maze-

runway situation.

Experiment 4. A study of the effect of dextran-induced sludge on

performance on a bar pressing task under a DRL schedule using a 20 sec.

minimum delay interval.

The results from these several experiments provided no evidence

for any significant behavioral effects attributable to the presence of

severe HmDx-induced sludge in the rabbit,

Experiment 1. The effects of dextran-induced intravascular aggregation
of red blood cells on food intake, water intake, body weight, home-cage
activity and exploratory behavior in the albino rabbit

Experiment 1 was intended primarily to test the hypothesis that

severe blood sludge is associated with a significant decrease in levels

of spontaneous activity. However, because of the scarcity of data rele-

vant to the question of the relationship between sludge and behavior,

the present experiment includes observations on several behaviors not

directly related to the hypothesis but which may nevertheless vary with

the level of sludge.

The possibility of major changes in behavior as a function of

duration of the sludge pointed to the need for observing behavior over


an extended period of time following treatment with the HmDx. In addi-

tion, while the severity of the experimentally induced sludge was known

to vary with time since treatment, no systematic observations had been

reported on the time course of the change. These two considerations

suggested that the study of the behavioral effects of sludge should be-

gin with a series of observations on both the dependent and the inde-

pendent variables covering a period of several days before and after the

sludge treatment.

Experiment 1 was designed to provide for the continuous monitoring

of a number of variables which were expected to vary as a function of

sludge. All observations were made over a 6 day period beginning 2 days

before treatment. The observations included (1) food intake, (2) water

intake, (3) body weight, (4) home-cage activity and (5) exploratory

behavior. Coincident observations were made on the condition of the

blood over the course of the 6 day period using the blood smear techni-

que for measurement of the severity of the sludge.


Subjects. Twenty young albino rabbits (obtained from Holsen-

beck's Rabbitry, Jacksonville, Florida), weighing between 1500 and 2500

gms, were tested for absence of red cell aggregation and assigned to

five replications of four animals each on the basis of body weight.

Within each replication the rabbits were assigned at random to four

groups: Group I, normal controls; Group II, saline-injected controls;

Group III, 77,000 molecular weight dextran-injected controls and Group IV,

the group receiving infusions of the sludge-inducing high molecular

dextran (HmDx).


Apparatus. The observations on several aspects of behavior

required the development of the following apparatus.

Activity cages. The four individual cages housing the

animals during testing were 10 in. wide by 12 in. high by LO in. long,

with sides, ends and the hinged tops made of 1/L in. hardware cloth and

the floors of 1/2 in. hardware cloth. Two standard glazed pottery food

and water containers (5 in. diameter by 3 in. deep) were placed side by

side at one end of each cage. The end of the cage was extended over the

dishes to allow free access to the dishes by the animals from the inside

and easy removal by the experimenter from the outside. The animals were

given Purina Rabbit Chow and tap water.

Home-cage activity measurements were obtained with a transistor-

ized (Silicon controlled switch, G. E. 2N58) photoelectric relay device

activating an Esterline Angus event marker recorder operating continu-

ously at a chart speed of 3 in. per hour. The photobeam controlling the

photosensitive resistor (Clairex, 1 megohm) crossed the cage 3 in.

above the level of the cage floor and 9 in. from the end of the cage

opposite to the food and water dishes. Interruption of the light beam

triggered the transistor and activated a recorder pen. Once the tran-

sistor had been triggered, the pen circuit became independent of the

photo-resistor and could be inactivated only by interruption of the pen

circuit itself. A narrow (5 in. wide) section of the cage floor located

between 10 and 15 in. from the end of the cage containing the food and

water dishes was hinged at its inner edge and made movable across a 0.5

mm arc at its outer edge to permit operation of a nearly silent micro-

switch which interrupted the recorder pen circuit when the section of


cage floor was depressed. The apparatus thus provided a virtually fool-

proof system for the measurement of home-cage activity: when the animal

moved to the rear of the cage it activated the pen circuit which then

remained on until the animal moved to the front of the cage and inter-

rupted the pen circuit by depressing the floor section and opening the


Each cage was suspended between 2 in. high "runners" over a lit-

ter pan spread with Pel-E-Cel. The cages were placed side by side in

individual compartments made from 1/4 in. plywood. These compartments

were 48 in. high by 60 in. long by 16 in. wide. A single door of 1/h in.

plywood permitted access to all of the cages at once.

Two levels of illumination were provided by a single 7 watt

frosted bulb and one 40 watt bulb at the top center of each compartment,

with the two levels controlled for all four compartments by a single s.p.

d.t. switch accessible from the outside.

"Square maze". The observations on exploratory behavior

were made in a four canpartment "square maze", a unit functionally simi-

lar to the circular maze used by Gwinn (1949) on rats. This unit was

33 in. square by l1 in. high and stood on 8 in. long legs. Its panels

were of 1/h in. plywood and painted gray. Each of the four compartments

was 16 in. square by lh in. high measured from the top of the removable

hardware cloth floor section. Square doorways, equipped with sliding

panel doors of 1/h in. plywood, were located between each two adjacent

compartments. These doorways were 8 in. square, and reached to within

2 1/2 in. of the tops of the floor sections. The sliding doors were so

placed that they were always behind the panel as the S moved in a clock-

wise direction from one compartment to the next. The wire cloth floor


sections were 3/L in. narrower than the compartment at each side and

were suspended only by their corners. Each compartment had a separate

hinged cover of 1/16 in. screening framed with 3/A in. by 3/A in. wood,

with a layer of cheesecloth above the screen to minimize the possibility

of the animal seeing beyond the limits of the apparatus. A single 7 1/2

watt bulb was placed at the center of each screened cover to provide

individual lighting for each compartment. Separate pull strings were

attached to the sliding doors between the compartments, but all doors

were connected to a single switch which controlled the operation of an

electric timer such that raising any or all doors activated the timer

while all doors had to be lowered before the timer circuit switch opened.

Rabbit holder. In this experiment as in all other exper-

iments described here, a V-shaped stand was used to hold the animal

while treatments were being administered or blood smears taken. This

stand was 18 in. long with sides 6 in. wide and made of 1/ in. p'ywoc:2.

A barrier was used at one end only, and the animal was placed with its

head directed toward the unobstructed end of the unit. 'Vhen this stand

was placed on a table with the open end slightly beyond the end of the

table a restless animal placed in the device would take one look at the

floor and retreat until it backed up to the barrier at the other end,

while the more timid one would back up to the barrier at once and stay

there. The unit was found to be extremely effective, and no major dif-

ficulties were experienced in treating or testing any of the large num-

ber of rabbits used in the several experiments reported here or in the

many exploratory studies which preceded the work discussed in this



Dextran treatments. The high molecular weight dextran

(2,000,000 molecular weight) was obtained from Pharmacia, Uppsala,

Sweden, and the 77,000 molecular weight dextran from Phannachem Corp.,

Bethlehem, Pennsylvania. All solutions were made fresh weekly as 10

per cent solutions in normal saline and autoclaved for h4 min. at 1180 C.

before use. All infusions were made at the rate of 10 cc per kg body

weight, thereby providing for a dextran dose of approximately 1 gm per

kg body weight. These solutions were administered into the marginal ear

vein through a 2 cm length of a 23 gauge needle inserted into the end of

a 40 cm length of .023 in. polyethylene tubing whose other end was at-

tached via a 23 gauge needle to a B & D two-way valve. This arrange-

ment permitted the use of a large (20 cc) syringe as a reservoir with a

smaller (1 cc to 5 cc) syringe providing the necessary infusion pressure

for the high viscosity dextran solutions. A constant temperature water

bath maintained the solutions and syringes at approximately the mean

body temperature of the rabbit (39.70 C.).

Procedure. At least four animals (one replication) were brought

into the laboratory at one time. Immediately upon arrival each animal

was weighed and a blood smear taken to determine the aggregation tenden-

cy of the blood. Any animal showing an abnormal condition of the blood

as defined by the condition of the blood smear was replaced at once.

The four animals chosen for each replication were then placed into indi-

vidual cages in the laboratory and provided with free access to food and

water during an adaptation period of at least three days.

At the end of the adaptation period, and beginning at 7 p. m.,

blood smears and body weight measurements were obtained, and the animals

were assigned at random to the four experimental cages within the


plywood compartments. These weighing and blood smears were repeated

at 12 hour intervals for the duration of the 6 .day observation period

which was standard for all replications. Daytime lighting (7 a.m. -

7 p.m.) for the compartments was provided by the hO watt lamps and the

7 p.m. to 7 a.m. lighting by the 7 watt lamps.

After a h8 hour standardization period the animals were assigned

at random to the four experimental conditions and the various infusions

were given immediately following the taking of the blood smears. Only

one infusion was given to any one animal within each replication. Obser-

vations were continued at 12 hour intervals until the end of the sixth

2h hour period for a total of h days of posttreatment observations.

At the end of this period the four animals were removed from their exper-

imental cages, a final series of blood smears and body weights taken and

the cages thoroughly cleaned and scrubbed preparatory to beginning the

next replication.

For observations of food and water intake during the 6 day period

covered by each replication, the food and water containers and their con-

tents were weighed at the beginning and at the end of each 12 hour period.

The differences between these two measures were taken as the amounts con-

sumed. Control was provided for evaporation loss. Fresh food and water

was supplied for each 12 hour interval.

Because of the number of different measures obtained during these

12 hour intervals, the handling procedures will be described in detail.

Measures of exploratory behavior were obtained as follows. All lights

except those in experimental cages and in the square maze were shut off.

The single large door of the experimental compartments was raised. The

animals were taken from their cages one at a time, carried to the square

maze and placed carefully into compartment 1. After a 10-sec. delay all

doors of the maze were opened at the same time and kept open for 60 sec.

During this time the animal was free to move from compartment to compart-

ment and his movements were traced out by E on diagrams of the maze.

These diagrams served as the basic data for the observations on explora-

tory behavior.

As each animal was tested in the square maze it was placed in a

temporary cage to await completion of the entire test series. Next,

body weights and blood smears were obtained, with the animals being han-

dled in the same order as previously. The animals were again returned

to their temporary cages while food and water containers were weighed

and refilled and the litter pans under the experimental cages cleaned

and washed (a.m. only). Following these operations, and at the end of

1/2 hour after the opening of the door to the experimental cage compart-

ments, the animals were returned to their respective cages to remain un-

disturbed for another 11 1/2 hours. The 1/2 hour period which was al-

lowed for the observations and handling was seldom exceeded except when

the various injections were given in addition to the regular handling.


The group mean results for the six different measures (including

sludge ratings) obtained in Experiment 1 are presented separately in

Tables 1 through 6 and are shown graphically in Figs. 1 to 6. Results

for individual animals for all observations are presented in Tables l1

through 19 in the Appendix.

Food intake. Group mean values for food intake (in gms) for


each of the four groups and for each 2h) hour interval of :he 6 day obser-

vation period are so awn in Table 1 an,! Fi z. 1. it is evident fro.

Fig. 1 that the largest drop in f-d intake on the first posttre:atnent

day occurs following treatment -.ith the sl.,' e-inducing : D(. ::Io.' r,

by the second day following treated -t the ,esn food intake for tis

, r -: rose to a level almost identical to that for the otl,hr .'.. o :reat~.

groups. Further, not all animals in te ':,iDx-treited 1 rou contrirtoi

to the initial posttreatrment fall in f oo intake levels. Food in~ .

for one of the animals in this group actually increased follow. ing; sl. -

i.-, (see Table l"), w]ile only two aniLals showed a consid rable decrease

at this time.

Table 1

h'ean Daily Food Intake (. ;) During 2
After Treatments with Five An:

;ys Before and for L Days
ials : Croup





: bilx


Day 1





Day 2



Day 3


After treatment

3 Day 5 r
- I-
1 111

In sur- ry, .ean daily

fell to below control r ,up I

veoo otae for thte

*vels for the f rst

Water intake. The rosaits fro7 ..1 1. of .. ta:e

by the fotau treatment .r.i, duri.: t'e 6 ,-day observation 3~-1 : crd

shown as mean values in Table 2 an are .ted rap ically .. Fg. 2.

1 ..r





Table 2

Mean Daily Water Intake (gms) During 2 Days Before and for h'Days
After Treatments with Five Animals per Group

Before treatment After treatment
group Day 1 Day 2 Day 3 Day 4 Day 5 Day 6

Normal 165 192 191 174 179 172

Saline 180 169 169 172 194 166

Dextran 228 243 243 274 232 231

HmDx 215 201 201 165 200 188

Individual results are found in Table 15. Fig. 2 shows that the

patterns of water intake for the dextran-treated groups are similar in

that both show a fall in water intake immediately following the injec-

tions and a rise by the second posttreatment day. However, evaluation

of these results is complicated by the fact that levels of water intake

exceeding 500 per cent of "normal" were reported for one of the animals

from the dextran control group. Therefore, the results for the HTDx-

treated animals should be compared only with the results for the normal

and saline groups. This latter comparison shows that the mean water in-

take for the naDx-treated animals returned to approximately normal values

by the second posttreatment day and continued to rise in linear fashion

to a maximum on the third posttreatment day.

Body weight. The results from the twice daily measurements of

body weight are presented in Table 3 as group mean values for each of

the 6 days of observations. The table also includes the average begin-

ning body weights for the four groups.





8 60
c DEXTRAN - - -
Before After
1 2 3 4 5 6
24 Hour Period

Figure 1. Mean Food Intake per 24 Hours for Each of Four Groups
of Albino Rabbits with Five Animals per Group.


200 -

Before After
12 3 4 5 6
24 Hour Period

Figure 2. Mean Water Intake per 24 Hours for Each of Four Groups
of Rabbits with Five Animals per Group.


Table 3

Mean Body Weight (gms) During 2 Days Before and for h Days After
Treatments with Five Animals per Group

Before treatment After treatment
group Start Day 1 Day 2 Day 3 Day 4 Day 5 Day 6

Normal 2058 2074 2085 2106 2122 21L6 2172

Saline 2121 2105 2128 2118 2169 2199 2228

Dextran 2151 2145 2159 2161 2182 2217 2258

fmDx 2153 2169 2189 2204 2222 2245 2284

These group results are shown graphically in Fig. 3 to facilitate

inter-group comparisons. Fig. 3 shows that following treatment the mean

body weight for the dextran-treated animals showed a fall in the rate of

weight increase from which the animals did not recover until the fourth

day after treatment. On the other hand, those animals treated with the

high molecular weight dextran showed no change in rate of weight increase

during the posttreatment period.

Home-cage activity. Home-cage activity was defined in terms of

the number of round trips made by the animal in its long narrow cage

during a given period of time. Measures of this behavior were obtained

directly from the chart of the Esterline Angus event marker recorder.

Results for individual animals for 12 hour and 21 hour intervals are

reported in Table 17 (Appendix). Group means values for successive

24 hour periods of the 6 day observation period are presented in Table 4

and Fig. L. As shown in Fig. 4, group mean differences are minimal for

the 24 hour period immediately prior to the treatments and for each of

the 4 days of the posttreatment period.


Table L

Mean Home-cage Activity (pacing) During 2 Days Before and far 4 Days
After Treatments with Five Animals per Group

Before treatment After treatment
group Day 1 Day 2 Day 3 Day h Day 5 Day 6

Normal 93 53 69 57 52 51

Saline 109 57 57 L8 51 43

Dextran 132 62 61 53 48 58

HmDx ll1 65 51 58 52 L7

Table 17 indicates that while two of the five HmDx-treated ani-

mals showed substantial decreases in level of activity for day 3 as

compared to day 2, two of the animals from the dextran control group

showed similar changes in activity levels. In summary, these results

offer no evidence for any changes in activity level following treatment

with HmDx.

Exploratory behavior. Results of observations made on explora-

tory behavior during the course of successive 1 min. exposures to the

square maze situation are presented in Table 5 as group mean values

shown separately for each of the 6 days of observation. These results

are presented graphically in Fig. 5. Individual scores for each testing

session are reported separately in Table 18. Fig. 5 reveals that the

saline group and the HmDx group showed almost identical performance for

each of the 6 days of observation, while both of these groups showed a

considerably higher level of activity in terms of mean number of com-

partments entered than did normal or dextran control animals.





Before Aft


24 Hour Period

Figure 3.


24 Hour Body Weights for Each of Four Groups of
Albino Rabbits During 6 Days of Observation
Treatment Given at the End of 2 Days.


80[ \.



3 4
24 Hour Period

Figure 4.

Mean Home-cage Activity (No. Round Trips) for Four
Groups of Five Albino Rabbits Before and After
Treatments During the 6. Day Observation Period.


- --- ------


2000 O




Table 5

Mean Exploratory Behavior During 2 Days Before and for 4 Days After the
Treatments with Five Animals per Group






IMean no. compartments

Before treatment

Day 1 Day 2

7.4 5.8

7.2 9.2

7.2 6.8

6.8 9.2

entered per two 1 min. test periods

After treatment

Day 3 Day 4 Day 5 Day 6

5.2 4.4 3.8 2.4

7.2 5.6 3.4 3.0

i.4 2.4 2.6 3.2

6.8 5.6 4.8 3.4

In spite of the size of the differences among the treatment means,

no formal analysis of the results was carried out because of the obvious-

ly large within group variability and the skewness of the distributions

of individual scores (see Table 18).

Blood sludge levels. Mean daily sludge ratings based on obser-

vations (blood smears) taken at 12 hour intervals during the experiment

are presented in Table 6 for each of the four treatment conditions.

Table 6

Mean Blood Sludge Levels During 2 Days Before and for 4 Days
After Treatments with Five Animals per Group






Blood smear ratings,

Before treatment

Day 1 Day 2

1.0 1.0

1.0 1.0

1.0 1.0

1.0 1.0

lowest (+1)

Day 3





to highest (+4) sludge levels

After treatment

Day 4 Day 5 Day 6

1.0 1.0 1.0

1.0 1.0 1.0

1.0 1.0 1.0

3.8 3.0 2.2

10 l

Sr__ c HMrDX

Before After

2 3 4 5 6
24 Hour Period

Figure 5.

Mean Number of "Square Maze" Compartments Entered by
Four Groups of Five Albino Rabbits Tested Daily at
12 Hour Intervals Over the 6 Day Period Beginning 2
Days Before Treatments.

--- .
, ~-~ c<

HMD ----



Before After

2 3 4 5 6
24 Hour Period

Mean Blood Sludge Levels During 2 Days Before and
for 4 Days After Treatments for Four Groups of Five



C 3


0 2



Figure 6.



Individual observations are reported in Table 19. These sludge

ratings indicate a high degree of reliability for the technique since

normal blood is reported for all animals and all conditions except for

the HmDx-treated animals following treatment.

Further, the course of recovery from the sludging effect of the

HmDx seems to be both gradual and uniform among all animals used in the

present study. The almost total absence of random variability in the

results obtained with the use of the blood smear technique for evalua-

ting the aggregation tendency of the blood might appear to indicate

the presence of bias in measurement. Accordingly it seems appropriate

to point out that the technique used here offered protection against

bias, first, through the use of a coded marking system which effectively

concealed the identities of the individual slides, and, second, through

randomizing the order in which the blood smears were rated. The results

of these observations indicate that the blood sludge induced through the

rapid infusion of 1 gm of HaDx remains severe for at least 2 days

following the infusion after which it begins to improve until by the

end of the fourth day after treatment most animals show only mild

clumping and rouleaux formation.


The results from the various observations made on the four groups

of rabbits before and after the several treatments suggest that, while a

modest and short-lived reduction in food and water intake may follow the

onset of severe red cell aggregation in the animal treated with high

molecular weight dextran, such reduction is not associated with a fall

in the rate of weight change. Further, the reduction could not be a


universal effect since not all animals showed it. The transient fall in

water intake was succeeded by an above normal increase in water consump-

tion. This phasic response in water intake may be related to the plasma

expanding characteristic of the dextran, and so the difference in rate

of recovery from the mild depression in rate of water intake may be a

function of the rate of loss of the dextran from the body. Thus the

effect upon water intake may be largely independent of the condition of

the red cell aggregation while the food intake should be presumed to

follow water intake and may also be independent of the sludge. Impres-

sive evidence for the independence of the changes in levels of water in-

take and severe red cell aggregation is to be found in the observation

that the pattern of water intake for the HmDx-treated group in the pre-

sent experiment appears to coincide with the changes in urinary excretion

levels which Gelin (1956) reported for the rabbit following the 1 gm/kg

dose of HmDx. As noted earlier in this report, Gelin found that the

urinary output on the first posttreatment day almost stopped altogether.

Then it began to rise gradually, until by the third day it had nearly

reached pretreatment levels.

The results from the observations on activity levels also indicate

the absence of any treatment effects associated with the blood sludge

per se. Spontaneous home-cage activity levels as determined by the num-

ber of times an animal traveled the length of its cage and returned was

found to be nearly equal for all treatment conditions for each day of


With respect to the measures of exploratory activity the results

were also negative in spite of the fact that consistent differences

were found to occur among the several treatment means over the course


of the observation period. The IlmDx and the saline groups showed com-

parable changes in levels of exploratory activity as a function of time,

and these levels were consistently higher than those for the other two

groups. However, these relationships among the group means developed

before the treatments were given and they showed little change during

the immediate posttreatment period. Also, the group results showed

large within group variability and abnormal distributions of individual

scores. While these two sets of observations, that of group mean

changes independent of treatments and that of the large random inter-

subject variability, may have discouraged a formal analysis of the re-

sults, they should permit the conclusion that if HmDx has any effect on

exploratory behavior that effect is not toward the suppression of such


Summary and conclusions

Observations were made on patterns of food intake, water intake,

weight changes, home-cage activity levels, exploratory behavior and

blood sludge levels in young male albino rabbits before and after treat-

ment with high molecular weight dextran to induce severe red blood cell

aggregation (sludge). Controls were provided for the effects of the

saline and of the dextran per se. The blood sludge resulting from the

1 gm/kg body weight dose of HmDx was found to remain severe for at least

2 days following treatment and to improve gradually thereafter, but no

behavioral consequences could be related to the presence of the severe



Experiment 2. The effects of sludge on shock-escape learning in the
albino rabbit

The generality of the negative findings from the previous experi-

ment was seriously restricted through the concentration on the study of

simple maintenance behaviors. This observation pointed to the need for

sampling a broader range of behaviors. The present experiment represents

a first step in the study of the effects of blood sludge on behaviors of

increasing complexity. This first step involved a simple shock-escape

conditioning situation in a beginning effort to define the effects of

severe blood sludge on learning.

Two problems were met in developing a testing situation for use

with the rabbit. First, the hair pads on the animals' feet offer a high

electrical resistance. This problem was resolved by using a high vol-

tage shocker. Second, the rabbit, or at least the strain used in the

present series of studies, shows a strong tendency to crouch and to lick

its feet in the presence of shock rather than to seek escape from the

situation. The results obtained from six replications of this experiment

clearly reflect the failure to fully resolve this problem.


Subjects. Twenty-four albino rabbits (M. E. Holsenbeck's Rab-

bitry, Jacksonville, Florida) were obtained and handled in the same

manner as described for Experiment 1. These rabbits provided six repli-

cations with animals matched for weight and aggregation tendency of the

blood within each replication but with treatments (normal, saline, 77k

dextran and HmDx) assigned at random within each replication. Body

weights of the animals ranged from about 1500 to 2500 grams.


Apparatus. The testing situation used here was the square maze

described in Experiment 1, Part II, but with several additional features

which were required because of the use of the shock. The "shocker" unit,

adapted from the one described by Licklider (1951), was a standard 12

volt automobile ignition system operated by a thyratron-controlled 110

volt electric motor to provide a pulse rate of approximately 15 per

second. The high voltage output of the system was estimated to be over

20,000 volts. A phototimer was used to provide a 2 sec. delay interval

between the opening of the compartment door and the activation of the

"shocker". The phototimer controlled a relay in the primary circuit of

the shocker unit. The system operated as follows: raising any door of

the square maze activated an electric stop-clock and the phototimer.

As the door remained open beyond the 2 sec. delay period, the photo-

timer relay circuit was energized, thereby causing the contacts of the

relay in the primary circuit of the shocker to be closed, thus activa-

ting the primary circuit of the shocker and producing the high voltage

pulses at the output of the ignition coil. This high voltage output was

fed through a high voltage L-contact rotary switch and separately to the

hardware cloth floor sections of each of the four compartments of the

square maze. The wiring used here was standard high tension automobile

ignition wire. Capacitance of the high voltage system was increased by

the addition of a h in. wide strip of 0.5 mm thick aluminum along the

two outer walls of each of the individual compartments of the maze. The

high voltage system was grounded to these strips.

The preparation and the composition of the various solutions used

in treating the animals and the techniques and materials used for the

infusions and blood smears were the same as for Experiment 1.


Procedure. Following the prehandling and an extended period of

adaptation to the laboratory as described under Experiment 1, the four

animals of each replication were assigned at random to the four treatment

conditions. In addition, the treatment and testing order for each repli-

cation was randomized and all handling during the experimental period

was performed in the same order.

On the day of testing each animal was removed from its cage,

weighed, tested for aggregation tendency of the blood (Smear A), given

the assigned treatment and returned to its cage. Then, 1 hour after

the beginning of the injection the animal was placed into compartment 1

of the test apparatus. (Only the lights in the apparatus were on at this

time and daylight was maximally excluded from the room.) Ten seconds

later one of the two doors of the compartment opened. The opening of

the door caused the switch controlling the stop clock and the phototimer

to be closed, thus initiating the 2 sec. delay period for the shock being

applied to the floor of the compartment containing the animal. Success-

ful escape from the shock required the animal's jumping through the open

door into the adjoining compartment. The door was released as the rab-

bit's hind legs touched the floor of the next compartment. Since lower-

ing the door also opened the switch controlling the timers, this act

also provided the end point for the measurement of response time.

Testing consisted of three series of three trials each, with 10

min. between series and 10 sec. between trials. By directing the shock

always to the compartment containing the animal and by raising the com-

partment doors successively in a counterclockwise direction the three

trials of each session moved the animal from compartment 1 to compart-

ment l. Finally, 10 sec. after the end of trial 3 the cover of


compartment 4 was raised and the animal gently lifted out and returned

to its cage. A total of three blood smears was obtained from each ani-

mal; one before treatment (Smear A), another before testing (Smear B)

and a third after testing was completed (Smear C). This approach per-

mitted an effective monitoring of the conditions of the blood throughout

each phase of the experiment. All blood smears were rated in a random

order using the method and the 4-point rating scale described under

Experiment 2, Part II.


Combined results for the six replications of four animals each are

presented as group mean values in Table 7 and 8. Table 7 shows the group

means for first trials for each of the three successive 3-trial series.

Group means for combined trials for each series are presented in Table 8.

Standard deviations are included. Results for individual animals are

fund in Table 20 (Appendix).

Table 7

Mean Response Latencies (in sec.) for First Trials from Three Successive
Series of Three Trials Each for Four Groups of Albino Rabbits in a
Shock-escape Situation with Six Animals per Group

First trials

Series 1 Series 2 Series 3
group Mean SD Mean SD Mean SD

Normal 62.8 61.0 10.9 6.9 10.h 6.0

Saline 99.6 84.1 39.9 73.2 34.9 62.9

Dextran 62.7 70.6 21.7 25.6 12.3 13.8

HmDx 44.2 64.L 12.2 15.8 8.8 7.3


Table 8

Mean Shock-escape Latencies (in sec.) for Combined Scores for Three
Successive Series of Learning Trials with Six Animals per Group

Series 1 Series 2 Series 3
group Mean SD Mean SD Mean SD

Normal 101.8 72.5 31. 13.5 29.8 10.0

Saline 130.6 86.4 91.6 1L6.1 109.9 188.2

Dextran 137.6 115.7 51.7 5L.5 41.7 50.0

HmDx 107.8 115.3 33.8 11.3 26.2 11.0

The mean values found in Tables 7 and 8 are presented graphically

in Figs. 7 and 8, respectively. Both reveal typical learning curves for

all treatment conditions with the exception of the saline group from

Fig. 8. The U-shape of the learning curve in this one instance is the

result of extremely long latencies shown by a single animal during the

third series of trials (see Table 20).

Comparisons among group performance levels as shown in Fig. 7 and

Fig. 8 show a close correspondence between changes in response latencies

as a function of trials for the normal and the HmnDx (sludged) group.

Even the ranges of the individual scores for both the first trials of

each series and for series totals are comparable. Intergroup comparisons

involving the saline and the dextran control groups are complicated by

the presence of one extremely high scoring animal in each of these two

groups. However, the results for the remainder of the animals from each

of these two control groups seem to compare favorably with the group

values shown for the normal and the HmDx-treated group.

Blood sludge ratings for each of the three blood smears (Smears

A, B and C) are shown separately for individual animals in Table 21






1 2 3

Mean Response Latencies for First Trials for Succes-
sive Series of Shock-escape Learning Trials




h .-

1 2 3

Figure 8.

Mean Total Response Latencies (in Sec.) for Four
Groups of Six Albino Rabbits for Each of Three
Successive Series of Three Trials Each in a
Shock-escape Learning Situation.

HMDX ---

Figure 7.







N -- - -

and are presented in summary form as group mean values in Table 9.

Table 9

Group Mean Blood Smear Ratings Before (Smear A) and After (Smear B)
Treatments and After Shock-escape Conditioning Trials (Smear C)
with Six Animals per Group

Sludge ratings: lowest (+1) to highest (+L)

Smear Normal Saline Dextran HmDx

A (pretreatment) 1.0 1.0 1.2 1.0

B (posttreatment) 1.0 1.0 1.3 h.0

C (posttesting) 1.0 1.0 1.2 4.0

Table 9 shows that the blood was essentially free of any aggre-

gation tendency in all groups prior to treatment and remained so follow-

ing treatment in all control groups, whereas severe aggregation resulted

in the HmDx-treated group following treatment. Of the 2h rabbits used

in the present experiment only the dextran-treated rabbit from replica-

tion 2 showed any pretreatment sludge tendency of the blood (sludge

level + 2) as measured by the blood smear technique, and this same level

of sludge tendency remained constant throughout the period of behavioral

observations. The dextran control animal from replication $ also showed

some sludging tendency of the blood (sludge level + 2), but only during

the immediate posttreatment period. These two animals also showed lower

response latencies for trial series 2 and 3 than did the other four ani-

mals of the same treatment group.


The high degree of correspondence between the means and variances

from each of the three series of test trials for the normal control group

and the HmDx-treated group showing severe sludge as determined by the

blood smear technique offers strong evidence that learning as measured

under the conditions of the present experiment is not seriously affected

by the sludge. Neither is there evidence of any major changes in reac-

tion times, particularly in the direction of increased latencies as would

be predicted by an hypothesis relating blood sludge to cerebral anoxia,

a condition which should be expected to promote a stuporous state (Levy,

1920). In fact, not only did the sludged animals show on their first

exposure to the shock-escape situation a mean response latency comparable

to that of the normals, but the shortest response latency on first expo-

sure to the test situation was obtained by an animal from the hmDx group.

Interestingly, of the six animals from the dextran control group, the

two best performers in terms of response latencies were the two which

showed mild blood sludge (sludge level +2) when tested by means of blood

smears immediately before exposure to the learning situation.

The importance of the large individual differences in performance

levels should be considered. First, as suggested by the preceding dis-

cussion, this extreme variability of the scores discouraged a formal

analysis of the data. Further, as indicated earlier, the common response

of the rabbit exposed to shock was that of crouching and retreating to a

corner of the compartment. Consequently, the latency measure used in the

present study may have been more of a measure of readiness to make a

particular response than a measure of readiness per se.

Summary and Conclusions

Twenty-four young male albino rabbits representing four separate

treatment conditions, a sludged group and three control groups, were


tested in a shock-escape learning situation beginning 1 hour after

treatments. No evidence was found that HmDx-induced sludge had any

effect on shock-escape learning.

Experiment 3. The effects of dextran-induced sludge on food reinforced
maze learning in the albino rabbit

While Experiment 2 was intended as a study of learning, the results

were inconclusive for several reasons. The large variability of the

individual scores could have concealed modest treatment effects. At the

same time the tendency of the animals to crouch instead of to run raised

the question of what the response latency measures signified. Finally,

the use of the shock-escape situation may have produced behavioral

changes which reflected the confounding of adaptation to the UCS with

instrumental conditioning.

The present study represents an attempt to obtain minimally con-

founded measures of learning under conditions of severe sludge. Food

instead of shock was used as the UCS. This change in procedure should

have minimized the probability of the animal making any response to the

CS (raising of the compartment doors) other than the one being measured.

Further, the use of food as the UCS should have assured the full depend-

ence of the measured response on the CS.


Subjects. A total of 24 male albino rabbits weighing between

1700 and 2300 gms was used for six separate replications of a random

replications experimental design with four treatment conditions: normal

control, saline control, dextran control and HmDx (severe sludge).


Apparatus. The present behavioral testing situation involved a

modification of the basic square maze unit used in Experiment 1. The

modifications included the removal of the shocker, the installation of

a small food cup at the outer corner of each compartment and the addi-

tion of a second electric timer.

Each food cup was 1 in. deep with tapered sides and a flat bottom

2 in. in diameter resting 1 in. above the canpartment floor. Food pel-

lets (250 mg Purina Rabbit Chow pellets) were introduced into these cups

through polyethylene tubes which extended from a common point at one side

of the unit to each of four short metal tubes which penetrated compart-

ment walls directly over the food cups. The second electric timer was

connected in parallel with the electric stop clock already in the circuit

operated by the compartment doors. A switch operated by the experimenter

permitted operation of either of the two clocks but only while the com-

partment doors were open.

Procedure. All animals for a single replication were obtained

at the same time, and all were tested for absence of significant blood

sludge before being accepted into the laboratory. Following a 5 day

minimum pretreatment adaptation period the animals were placed on a food

deprivation schedule consisting of one 16 hour period followed by two

periods of 70 hours each, with treatments and behavioral testing begin-

ning at the end of the second 70 hour period. Drinking water was freely

available at all times except during treatment or testing.

The procedure followed on the test day was as follows. All ani-

mals were removed from their home cages, weighed, tested for absence of

blood sludge by means of blood smears and returned to their cages.

Four animals, matched by weight and condition of the blood, were randomly


assigned to the four treatment conditions and to separately randomized

testing orders. In accordance with the treatment and testing schedule

each of the four animals was removed from its cage, a fresh blood smear

'taken (smear A), the appropriate treatment administered, and the animal

again returned to its cage. Fifty-eight minutes after the beginning of

the treatment the animal was again removed from its cage and another

blood smear (smear B) taken, but this time the animal was placed immedi-

ately into the square maze.

A standardized testing procedure was used for every series of

trials. All the lights in the laboratory were shut off 1 min. before

testing was begun. The animal was always placed into compartment 1 of

the maze. Following a 10 sec. delay interval, three of the four com-

partment doors were raised simultaneously to open a path from compart-

ment 1 to compartment h in a counterclockwise direction and at the same

time to start the electric timer #1. As soon as the animal hopped

through the first doorway and its hind feet touched the floor (compart-

ment 2) the experimenter switched over to timer #2. This procedure

provided separate measures of start latency and running time.

The food reinforcement was introduced into the appropriate food

cup before each trial. For trial 1 the 250 mg Purina Rabbit Chow pellet

was present in the food cup of compartment 4 before the animal was intro-

duced to the apparatus. For trial 2 the food reinforcement was released

through the plastic tubing into the appropriate food cup (compartment 3)

at the moment that the doors were lowered at the end of trial 1. This

method prevented the animal from orienting to the correct food cup on

the basis of the sound of the pellet striking the food cup. Timing of the

30 sec. interval was begun only after the animal found the pellet in

the food cup.

The learning trials consisted of three series of five food

reinforced trials followed by one set of five extinction (non-reinforced)

trials. The first of these series began 1 hour after treatment and

each successive test series beganat /2 hour intervals. No intertrial

handling was necessary during the five trials of any one series. The

animals simply kept circling in a counterclockwise direction, with the

goal box for the first trial of any series becoming the starting box of

the next trial such that the starting box moved clockwise until, with

the final trial (trial 5) of a series, compartment #1 was again the start-

ing box. A final blood smear (smear C) was taken upon completion of the

four successive test series.


Group mean values obtained from the measurements of starting

latency and running time are presented in Tables 10 and 11. The means

shown in Table 10 are for first trials only and are shown separately for

each of the three series of reinforced trials and for the single extinc-

tion series. These results are presented in Fig. 9. Data for individual

animals are found in Table 22 (Appendix). Similarly, group means for

totals for the five trials from each of the several series of trials are

presented in Table 11 and in Fig. 10. However, the values shown for in-

dividual animals in Table 23 represent sums for the five trials from each

series rather than individual trial scores. Standard deviations are

reported in all tables.

Some of the values shown in Tables 10 and 11 are based on the

results from less than 6 animals. Such discrepancies occurred because

testing was discontinued within any given series for any animal which


exceeded 500 sec. on either of the two measures of performance while

no series total scores were included in the results for individual ani-

mals unless all five trials of the series were completed. Because some

animals completed some but not all of a series of trials, fewer omissions

are found in the results for first trials as shown in Table 10 than in

Table 11 which reports series totals. Further, it should be noted that

the results for first trials as shown in Table 10 and Fig. 9 permit mean-

ingful comparisons among all four series of observations since both re-

sponse latency and running time for this first trial of the extinction

series should have been independent of the conditions of the reward.

Table 10

Mean Response Latencies and Running Times (in sec.) for First Trials From
Each of Three 5-Trial Food Reinforced Test Series and One Extinction
Series (4) for Four Groups of Rabbits with Six Animals per Group


Normal III


Saline Mff


Dextran M


HmDx M1












response latency

2 3 4

10.3 10.5* 10.8*

7.1 8.7 6.9

8.7 8.2 10.4

5.5 L.8 9.5

19.2 13.6 5.4*

14.1 11.4 2.8

17.6 8.4 17.1

5.6 5.3 21.0

Mean running time

1 2 3

91.9 73.1 28.0*

73.6 128.7 30.8

65.3 32.2 42.7

36.1 51.1 74.5

59.5 16.3 35.7

33.2 11.9 38.0

79.6 32.8 25.1

42.9 35.1 37.7










Table 11

Mean Total Response Latencies and Running Times (in sec.) for Each of
Three Successive 5-Trial Series of Food Reinforced Learning Trials
and One Extinction Series (4) for Four Groups of Six Albino
Rabbits Tested in a 4-Compartment "Square Maze"


Normal M


Saline M


Dextran M


HmDx M












response latency

2 3 2

34.5* 2.1" 47.4*

16.0 19.4 18.8

14.6 35.2 46.1

23.3 13.9 33.2

53.7 51.3 33.6*

22.1 28.3 14.5

66.8 47.4 40.8*

38.4 25.2 26.9










Mean running time

2 3

9 76.5* 13L.1*

8 35.3 79.5

5 135.9 133.8

8 152.0 126.5

3 120.8 103.0

5 61.7 93.5

0 121.2 119.2

6 77.9 97.2

* N= 5

N= h

The within groups variability appears to be relatively high for

both sets of measures as indicated by the standard deviations shown in

Tables 10 and 11. Some of this variability should be related to the

separate replications because of the use of the random replications

design. However, most of the variability seems to be uncontrolled. For

example, the two obviously deviant mean scores occurring for the sludged

group as shown in Fig. 10 are readily accounted for on the basis of sin-

gle extreme scores. If the single high latency score (921.1) for the

HmDx-treated animal from replication 1 was ignored, then the mean for

the sludged group for the sum of all trials of series 1 would be 92.4













1 2 3 IS
Test Series

\ \'


1 2 3 li
Test Series

Mean Group Results Using First Trials for Measures of
Response Latency and Running Time from Four Groups of
Six Albino Rabbits Receiving Three Series of Food
Reinforced Training Trials and One Extinction Series.






1 2 3 4I
Test Series

1 2 3 I
Test Series

Figure 10.

Mean Group Results for Measures of Response Latency
and Running Time Obtained from Four Groups of Six
Albino Rabbits Tested in a "Square Maze" with Three
Series of Positively Reinforced Learning Trials and
One Extinction Series.




Figure 9.



,-s ---


0 100




instead of 231.h (see Table 23). This adjusted value would be almost

exactly equal to the mean for the dextran control group (91.6) which is

the lowest of the four for this series. In similar fashion the adjust-

ment of the mean running time value for the dextran control group for

series 1 by the omission of a single extremely high score would serve to

equate the results for the sludge group and the dextran control groups.

No other single deviant mean scores for the sludged group are seen in

Figs. 9 and 10. Neither is there evidence of trend differences. Be-

cause of the several extreme scores together with the absence of any

major differences among the means for the HmDx group and the various con-

trol groups, no formal statistical analysis was performed on the data.

Observations on the condition of the blood in all animals used in

the experiment are reported as group mean values in Table 12 while Table

25 also shows the ratings for individual animals.

Table 12

Mean Blood Sludge Levels Before (Smear A) and After Treatments (Smear B)
and After Four Series of Training Trials (C) in the "Square Maze".
with Six Animals per Group

Smear Time Normal Saline Dextran HmDx

A Pretreatment 1.0 1.2 1.0 1.2

B Posttreatment 1.2 1.2 1.7 h.O

C Posttesting 1.2 1.2 1.0 h.O

These observations show that even the normal animals had mildly

sludged blood and that many of the animals (67 per cent) treated with

the 77,000 molecular weight dextran had a temporary increase in blood

sludge levels in the immediate posttreatment period which in every

instance returned to normal by the time behavioral testing was completed.

The results from the measurement of food intake during the 2 -hour

feeding period following behavioral testing are shown individually in

Table 2h. The mean food intake for the several groups during this

2-hour period was found to be as follows: normal control group, 21.3 gms

( .0), saline control group, 21.5 gns ( 5.2), dextran control group,

22.2 gms ( h.0) and HmDx group, 26.3 gms ( 6.7). Analysis of variance

on these results yielded a nonsignificant F (p > 0.05).


The testing situation used in the present experiment was intended

to provide a standardized learning task as a background against which to

measure the effects of sludge on learning. The results for the several

treatment groups presented in Figs. 9 and 10 reveal consistent changes

in behavior as a function of trials whether that behavior is measured in

terms of response latency or running tire. The situation therefore

should have been appropriate for the study of the behavioral effects of

sludge. While the blood sludge level in the severely sludged animals

appears to have been effectively controlled by means of the HmDx infu-

sions, blood smears obtained from the dextran control group show that

these animals are more sensitive to the effects of the 77k molecular

weight dextran than the animals from Experiment 1. This increased sensi-

tivity of the blood to the 77k dextran may have been related to the pro-

longed period of food restriction which preceded the treatment. However,

this increased responsiveness to the control dextran infusions should

not have influenced significantly the results from the experiment since

(1) no animal from the dextran control group showed sludge levels great-

er than that found to occur following 20 per cent of the standard dose


of HmDx (sludge level #2) and since not even the standard dose of HrDx

seems to have any effect upon behavior under the conditions of the

present experiment.

Food intake levels during the immediate posbtesting period were

not affected by the condition of severe blood sludge induced by the

infusions of HmDx. This latter finding might appear to be inconsistent

with the results obtained in Experiment 1 where mean food intake appeared

to be mildly depressed following the lnDx treatment. However, the dif-

ferences in the conditions under w:ich the respective observations were

made may account for the differences in results. The observations in

Experiment 1 were made on animals with continuous and free access to

food, and measurements of food intake were made for 12 hour intervals

rather than the 2 hour period used in the present study. It may be sug-

gested, therefore, that in the present experiment the HmDx associated

anuria did not have ti-e to affect the measures of food intake during

the short time period intervening between treatment and the measurement

of food intake. Actually, the present finding of a higher mean level of

food intake for the HmDx-treated group might be regarded as strong sup-

porting evidence for the conclusion reached in Experiment 1 that rmDx-

induced blood sludge per se has no significant effect on food intake.

Summary and conclusions

The effect of severe HmDx-induced blood sludge on behavior was

studied in rabbits using a food reinforced maze-runway learning problem.

Separate measures were obtained for starting latency and runrni.., time.

Neither measure provided any evidence for an effect due to the presence

of the sludge.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs