Title: UBVRI photometry of variable red dwarf emission objects
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098174/00001
 Material Information
Title: UBVRI photometry of variable red dwarf emission objects
Physical Description: v, 146 leaves. : illus. ; 28 cm.
Language: English
Creator: Martins, Donald Henry, 1945-
Publication Date: 1974
Copyright Date: 1974
Subject: Red dwarf stars   ( lcsh )
Photometry   ( lcsh )
Astronomy thesis Ph. D   ( lcsh )
Dissertations, Academic -- Astronomy -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis -- University of Florida.
Bibliography: Bibliography: leaves 143-144.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098174
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 - 000580925
oclc - 14100423
notis - ADA9030


This item has the following downloads:

ubvriphotometryo00mart ( PDF )

Full Text








Abstract. . . . . . .... . . . . iv


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

General Characteristics of the
Red Dwarfs . . . . . . . ... 2

Points of Interest . .... . .. . . 10

2. SMALL SCALE VARIABILITY. . . . . ... 13

History. . . . .. . . . . 13

The Problem. . . . ... . . 15

LP101-15--A Special Case . . . ... .17


Existing Equipment . . .... . . 20

New Equipment Required . . . . .. 21

4. A NEW PHOTOMETER . . . . . . . 27

The Offset Guider. . . . . . ... 27

Comparison Star Monitoring . . . ... 31

Multi-Channel Photometer Head. . . .. 32

Photomultiplier Chambers . . . ... 38

Data Acquisition . . .... . . . 44


Chapter Page


Kitt Peak National Observatory
Existing Equipment . . .. . . 47

RI Observations from
Lowell Observatory . . .... . . 55

Rosemary Hill Observatory. . . . ... 61


Pulse Counting . . . .... . . 66

Digital. . . . . . . . . . 74

7. TABULATED DATA . . . . . . . . 77

8. ANALYSIS . . . . . . . . . 99

LP101-15 . . . . . .... . . 105

BY Draconis, GT Pegasus, FF Andromeda. . 116


10. CONCLUSION . . . . . .... . . 138

BIBLIOGRAPHY. . . . . ... . . . . 143

BIOGRAPHICAL SKETCH . . . .... . . . 145

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Donald Henry Martins

March, 1974

Chairman: Frank Bradshaw Wood
Major Department: Astronomy

A review of the history and nature of small scale

variability of red dwarf emission line stars is presented.

This discussion is accompanied by a description of equipment

developed for the purpose of expediting data acquisition

for photoelectric photometry of faint stars.

Data obtained at the Kitt Peak National Observatory,

the Lowell Observatory, and the Rosemary Hill Observatory of

the University of Florida are tabulated and analyzed to

determine their implications with respect to the known vari-

ability of four stars. BY Draconis, GT Pegasus, FF Androm-

eda, and LP101-15 (an eclipsing binary) are studied.

LP101-15 and BY Draconis were observed on the Johnson UBVRI

system at Kitt Peak while BY Draconis, GT Pegasus, and

FF Andromeda were observed on the Johnson RI system at the

Lowell Observatory. UBV"R" observations were carried out

on all four objects at Rosemary Hill Observatory since the

spring of 1972.

Analysis of the data centers on discussions relative

to the proposed "spot model" to account for the observed

variations in V, R, B-V, V-R, and R-I. After allowance is

made for the influence of flare activity (found to be very

important), the spot model appears to be adequate in its

predictions when compared with observations.

A period of 0.63398 day is supported for LP101-15,

and a possible secondary minimum is proposed (this secondary

eclipse would be less than 0.05 magnitude deep in V),

assuming the above period to be correct. This star also

appears to exhibit small scale variability with a period

of 0.63398 day implying synchronous rotation and revolution.

Flare activity is seen to be very important on this star as


Continued observation of these stars and similar

objects is suggested, since further data are essential for

complete understanding of their behavior.

Chapter 1


It is a somewhat ironic fact in observational astro-

nomy that the most populous known class of stars within this

galaxy is also probably one of the least well understood and

observed. Generally referred to as red dwarfs, this class

embraces a tremendous variety of very different objects

which are usually further subclassified. Only two charac-

teristics are truly common to all members: low mass (typi-

cally less than 0.5 solar masses) and low surface temper-

ature (typically 2500-35000K).

Several factors are directly responsible for the

lack of knowledge with respect to these stars, and they

will be discussed in greater detail in this analysis

together with suggestions for observational techniques and

equipment to relieve the situation. Very briefly, though,

the primary factor is their intrinsic faintness which makes

it very difficult to obtain reliable data from a large


The purpose of this investigation has in reality

been two-fold. First, a continuing program of monitoring

some selected red dwarf emission stars was initiated to

provide hitherto unavailable information as well as to

expand upon that data already known. Second, recognizing


the instrumental limitations incumbent upon this type of

work, a complete multi-channel photoelectric photometer

system was designed and constructed for the primary purpose

of making the most efficient possible use of available

telescope time.

What follows is a detailed account of the current

status of the work, a presentation of those conclusions

possible at this time, and suggestions for future research.

General Characteristics
of the Red Dwarfs

As stated above, the red dwarfs are the most

populous group of stars known within our galaxy. They

also probably comprise the bulk of the mass of the system,

even though the individual masses are relatively quite

small. The range in mass is considerable, with the upper

and lower limits being indefinite to a large degree.

Unfortunately, the lower limit is the least well known,

and is intimately connected with the question of the least

amount of mass required in order to have sufficient

gravitational energy released upon collapse to generate

nuclear reactions in the core and sustain them over

astronomical time scales. At the present time, this lower

limit appears to be somewhat greater than the mass of


Table 1.1 lists pertinent data for some of the best

known red dwarfs as given by Limber (1957) based upon work

by Lippincott, Hall, Johnson, Morgan, Kuiper, Joy,

Van de Kamp, and Baade. All of these stars are members of

binary or multiple star systems and near the sun, by

virtue of which the information can be determined.

Table 1.1

Examples of Physical Parameters of
Red Dwarf Stars

Star log(M/M ) Mv Te(K) log(R/R ) Mbol

Kr 60A -0.572 11.84 3100 -0.460 9.60
Kr 60B -0.788 13.39 2900 -0.600 10.58
Ross 614B -1.116 16.8 2700 -1.02 13.0
Castor C -0.194 9.14 3460 -0.208 7.84

Limber (1957)

Also of interest here is a survey by Van de Kamp

(1971) in which all known stars within 5.2 parsecs of the

sun have been reviewed. Table 1.2 gives a breakdown of the

frequency of spectral type and Table 1.3 gives the repre-

sentative mass densities of the various spectral types.

The point to be made here is that with the large

number of red dwarfs known, even just in the solar

neighborhood, accurate information is available only for

the four objects listed above. It is obvious that more

data are badly needed in order to understand the processes

which operate in their interiors and atmospheres. For

example, while the proton-proton cycle is probably the

dominant energy production process in the core, other

processes are possible such as transmutation of elements

(boron, lithium, deuterium, or beryllium) or Kelvin

contraction with a subsequent release of gravitational


Table 1.2

Frequency of Spectra

Spectrum Number of Stars

A 2
F 1
G 3
K 8
M 40
White Dwarfs 6

Van de Kamp (1971)

Table 1.3

Mass Densities by Spectral Type

Spectrum Mass Density
(10 M per pc3)

0-B 0.9
A 1
F 3
dG 4
dK 9
dM 29
gG 0.8
gK 0.1
gM 0.01
White Dwarfs 8
Subdwarfs 1.5

Allen (1963)

The basic work on the interiors

back primarily to Osterbrock (1953), who

of red dwarfs goes

attempted to

produce a model to encompass the known data. To a certain

extent he was successful, by postulating extensive convec-

tive envelopes encompassing radiative cores. When applied

to stars between dKl to dMO, the theory fits observation

very well, but difficulties begin to arise with the later

spectral types. For instance, Osterbrock's model fails

for Kruger 60A at around dM4. Limber (1957) expanded

upon Osterbrock's model to allow completely convective

models in which electron degeneracy effects were taken

into account, and obtained good correlation with known

data for Kruger 60A (dM4) and Kruger 60B (dM6).

One fact stands out in these theories--the need

for well determined masses, radii, and luminosities for

late M type dwarfs. The latter statement indicates one

of the eventual goals of the current and future research

to be discussed in this paper. It is not a simple, short

term effort, however, because precise values of the

parallax of the systems under study would be needed for

truly meaningful results, especially for cases other than

double line spectroscopic binaries which are also eclipsing

binaries. Reliable observations for determining parallaxes

of promising objects are indeed being made on a continuing

basis at several observatories, with the selection being

made primarily from proper motion survey catalogs. At

least one of the objects under study in this paper

(LP101-15) is on the observing programs of Yerkes, Lick,

and Allegheny, as well as the Naval Observatory, but the

results so far are really only preliminary.

The proper motion surveys mentioned above are

really the best means of initial detection of nearby

objects, since probably all, or most of the brighter

nearby objects have already been observed. Any remaining

nearby stars will probably be quite faint and difficult

to observe except with large telescopes and sophisticated

equipment. The probability that a very distant star

will have proper motion equal in magnitude to that of

a nearby object is sufficiently small to warrant a concen-

tration on those very red stars which do exhibit an

extraordinarily large motion.

Spectroscopic surveys are very useful for detecting

specific objects such as late type red dwarfs; however,

these surveys are typically limited to visual magnitude

around +13 to +14, and since the absolute magnitudes of

the later types (around dM6) are also +13 to +14, it is

obvious that they will only be accessible out to around

10 parsecs. This of course is a severe distance limitation,

so the faint star proper motion surveys are really the

only way to detect large numbers of the slightly more

distant stars.

The above discussion serves to a certain extent to

illustrate the difficulties associated with obtaining

sufficient data for meaningful statistics in attempts to

establish population differences, galactic distribution,

mass density, etc. Most of these problems can be handled

by photographic techniques and astrometry, with some

assistance from spectroscopy with large telescopes. None-

theless, there is a great deal which can be contributed

by modest size telescopes if they are outfitted with

proper instrumentation. Photoelectric multi-band photom-

etry (UBVRI, for example) and programs for careful moni-

toring of brightness variations are both useful approaches

adaptable to smaller telescopes. They would permit

establishing good light curves for any eclipsing binaries

which might be discovered and would yield photometric

data useful for classification of stars observed. Well

observed eclipsing systems hold promise for divulging a

great amount of information applicable to the interiors

of these stars and should be vigorously studied.

Sufficient information is available at present to

allow the establishment of population differences among

the red dwarfs, as well as at least a probable distribution

in space around the sun. While the distribution can be well

established for main sequence A stars or G and K giants

for example, out to as far as two kiloparsecs, the limit for

the dwarf K stars is near 200 parsecs, with the limit for

the dwarf M objects being even less.

McCuskey (1965) gives a review of current knowledge

of spatial distribution of common stars in the galactic

plane, summarizing his results in several graphs. He

illustrates the trend of space density for dG to dK stars

as averaged over all galactic longitudes from 1I = 440 to

232. Apparently, out to the limit of the data, there is

no real deviance from a uniform distribution, though

regions of different density do exist.

Since the red dwarfs are so faint, it is apparent

that up to a point, the number of objects seen should

increase with decreasing apparent brightness. Figure 1.1

illustrates the observational evidence supporting that


In a single spectral class, a variation in relative

density with galactic latitude can also be seen, due

primarily to a concentration of giants near the galactic


The general distribution of several types of stars

in Z is shown by Elvius (1965). He estimates that for

dG8 stars around absolute magnitude +6, the density drops

to approximately 1/20 its value in the solar neighborhood

within one kiloparsec of the galactic plane. For absolute

magnitude +8, it drops to about 1/6. Data for fainter

stars are not complete enough for conclusive statements,

but the trend probably continues.

Finally, there is one other interesting feature

of the dwarf M stars which is relevant to this paper.

Delhaye (1953) discovered a basic difference between the

kinematic properties of normal M dwarfs and the M dwarfs

which exhibit emission lines in their spectra. Namely, the

Me stars exhibit a considerably smaller solar motion and

velocity dispersion. Delhaye (1965) finds a mean velocity


50 G STARS .

.% o
Dworfs 50

25 -

"-. K STARS / "

4 8 12
Apparent Visual Magnitude
Figure 1.1. Using the data of Becker (1942), the
percentage of dwarf stars in spectral classes G and K
is plotted versus apparent visual magnitude. Current
data support these results. This plot has been made
similar to Figure 15 in McCuskey (1965).

with respect to the sun of almost 11 km/sec 2 (probable

error) with a standard deviation of 18 km/sec for the

latter, while the corresponding values for the M stars

are 22 km/sec 1 (probable error) and 30 km/sec. Mihalas

(1968) gives a spread in H velocity of approximately

50 km/sec and a spread on 6 velocity nearly 30 km/sec

for the Me objects, with 86 km/sec in H, +40 to -70 km/sec

in 9 for the M stars. The dMe stars then are interpreted

to represent pure Population I objects, while the dM

stars are probably mixed Populations I and II. Apparently,

among F, G, and K stars, as well as M, strong line stars

are members of Population I while weak line objects repre-

sent Population II.

Points of Interest

As pointed out above, the dMe stars are considered

to be Population I objects still forming a relatively

coherent kinematic group. This of course implies that

they are young, perhaps not quite evolved onto the main

sequence, which makes them interesting subjects for research

on several points.

First, any new data on masses, luminosities, and

radii (plus chemical composition), give a direct check for

pre-main sequence evolutionary tracks.

Secondly, many of these dMe and dKe stars are

known to be flare stars; that is, they sporadically show

increases in brightness of varying intensity (typically the

range is from a few tenths of a magnitude up to two or

three magnitudes). Clearly there are processes at work

here which are not totally understood and need to be

studied--with a possible result that current theories of

energy generation in stellar atmospheres or in the outer

layers of these stars would necessarily be modified.

Futhermore, many of these same objects exhibit

small scale variations in brightness which are very diffi-

cult to explain, whether by reason of their nonperiodicity,

quasi-periodicity, amplitude variations, peculiarly

shaped light curves, or the like. In order to completely

perceive the explanations for any of these effects a

great deal of observing will have to be done beyond that

already completed. This paper, then, constitutes a summary

of the work done to date on these problems and a small

contribution of related data.

One very interesting observation may be noted here.

Upon a review of the Lowell Proper Motion Survey, the

frequency of occurrence of red dwarf-white dwarf proper

motion pairs and definite red dwarf--white dwarf binary

systems is seen to be quite high for the known proper motion

pairs. The probable explanation is that most of these

pairs are widely separated binary or multiple systems. The

existence of such systems is not unlikely, and Chandrasekhar

has analyzed the kinematics of such an arrangement, finding

it stable for astronomical time spans. The fact that one

member of these pairs is usually a white dwarf (red-red

dwarf proper motion pairs also occur) is very interesting

from an evolutionary standpoint and should be explored. As

they evolved, the present white dwarfs could have exerted

considerable influence on their red dwarf companions,

assuming such pairs to be true systems.

The subjects of the current research are one

probable eclipsing binary system whose primary is a dM3.5e

star (with a white dwarf proper motion companion) as well

as a small scale flare star, and several dMe and dKe stars.

The latter are thought to possess large scale cool areas

on their surfaces to account for their quasi-periodic,

variable amplitude brightness changes. Each object and

type of variability will be discussed in complete detail

later, but they serve here to illustrate the possibilities

for research in this rewarding area.

Chapter 2


As indicated above, the red dwarf stars exhibit a

wide range of types of variability. In general, from star

to star it is possible to observe slow, quasi-periodic vari-

ations up to flares of high intensity. Often one star

will display almost all of the known effects with large

scale flare activity superimposed upon the less violent


One interesting fact is the observation that appar-

ently the emission line stars are the ones which exhibit

intrinsic variability. Being an emission line star is not

a guarantee of variability, however. In general, probably

the great majority of the red dwarf emission stars are

peculiar in some fashion--only requiring sufficient

observation for the peculiarity to be detected.

In this chapter, however, discussion will be limited

to the area of small scale variability of the emisssion line

stars, including a brief discourse on flare activity as

exhibited by some members of the group under study in this



Small scale variability was first observed in these

objects more or less by accident while they were being

watched for flare activity, except for HD 16157 (CC Eri)

which was first observed spectroscopically as a binary by

Evans (1959). Subsequent photoelectric work showed that

CC Eri is also a light variable of small amplitude (which

may change from season to season) not readily explainable

by a binary hypothesis.

Popper (1953) has observed H-Hse lines as well as

the H and K lines of calcium in emission in the spectrum of

HD 234677 (BY Dra), and noted the hydrogen lines to be

varying in intensity on some plates. This led Chugainov

(1960) and Masani (1954) to begin observing the star for

flare activity, but none was actually seen photoelectrically

at that time.

In 1965 Chugainov found a periodic variation in

BY Draconis with a peak to peak amplitude of approximately

0.2 magnitude in yellow light with a period of roughly

3.826 days. Previous observations in 1954 and 1960 had

yielded no evidence of any variation whatsoever, and

furthermore, observations later in 1965 and in other years

indicate the presence of considerable phase shift as well

as changes in amplitude.

Nor are these the only known examples of this type

of behavior. Krzeminski (1969), and Torres and Mello (1973)

have published lists of the established variables of this

species which give all pertinent data, so they will not be

repeated here. Of these stars, three have been chosen for

study in this program, as well as another red dwarf emission

object considered to be an eclipsing binary.

The Problem

The light curves exhibited by the three stars

listed in the next paragraph (and by those stars specified

in the above mentioned references) are of a unique

character and very difficult to explain. Therein lies one

of the two major aspects of the current research.

As indicated, BY Draconis has been observed to

exhibit a relatively large amplitude variation (02 in 1965).

GT Pegasus (AC+31070565) and FF Andromeda (BD+340106) on

the other hand have only shown very slight changes in their

brightness (except for some relatively large scale flare

activity) on the order of 0.06 magnitude in V, but the

appearance of the curves is still the same. All of these

stars fluctuate in sinusoidal fashion, reminiscent of

a "contact" eclipsing binary light curve. However, the

available spectroscopic data indicate a separation of

components on the order of ten times their respective

diameters. This argues strongly against the eclipse

hypothesis, as do the changing phase and amplitude of the


The problem then is to account for the existence of

these fluctuations, the change in phase from epoch to

epoch, and the change in amplitude of the variation, as

well as the shape of the light curve itself.

Two possible explanations have been proposed for

the behavior just described. One suggested by Kron (1950),

Chugainov (1966), and developed by Krzeminski (1969),

postulates the presence of "spots" on the surface (regions

of unequal surface brightness). A second explanation

proposed by Evans (1971) suggests the possibility of "grey"

material falling through the inner Lagrangian point

between the components of what now must be binary systems.

Several difficulties arise in the application of

either of these models and more observations are needed to

resolve the situation. For example, either hot or cold

spots may account for the observed variation in any one

color, or band, and to distinguish between them one needs

color indices, U-B, B-V, V-R, R-I, for instance, in addition

to brightness measurements.

The primary difficulty with the "grey veil" model

proposed by Evans appears to be one of stability of material

at the inner Lagrangian point, as computations by Torres

and Mello (1973) indicate. As they state, non-gravitational

forces must be present to provide sufficient material for

sufficient lengths of time, at least when applied to CC Eri

and YY Gem.

It then appears that the spot model is the most

likely explanation, but it encounters even more trouble

than indicated above, in the ultraviolet excess observed

at minimum light. This can be at least partially accounted

for in terms of the flaring activity also present, so it

does not completely rule out application of the spot model.

In order to test this model more fully, then, the

present observing program was initiated in 1971 in an

attempt to provide some of the necessary observations. It

was obvious at the beginning that a longer baseline was

needed (at least to R and I wavelengths), especially for

these very red stars. Torres and Mello (1973) also indi-

cated the desirability of observations in these bands.

From the necessity of this extended range came the impetus

toward multi-channel design for the new instrumentation.

LP101-15--A Special Case

LP101-15 and its common proper motion companion

LP101-16 (a white dwarf) first gained attention as a can-

didate for "pygmy" status according to Zwicky (1963a, b,

1965, 1966). Eggen and Sandage (1966) took considerable

issue with this view after re-examining Zwicky's data and

contributing much of their own. They finally showed con-

clusively that LP101-15 is a normal dwarf M emission star

of proper luminosity for its spectral class based upon the

accepted (M ,R-I) main sequence.

In the course of their investigation, however, they

discovered that LP101-15 is an eclipsing binary, with a

period of 0.63398 day or 1.26796 days. The ambiguity arose

because they could not state with certainty whether or

not a secondary eclipse occurs.

If this star can be sufficiently well observed to

permit the construction of a good light curve, it may yield

a further calibration of the physical parameters encountered

in these low mass objects. Since it is even later in

spectral class than YY Gem (Joy has estimated LP101-15 at

between dM3e and dM4e), such data would be most useful.

The second aspect of this project was therefore

devoted to obtaining as much data as possible on this star

(excluding LP101-16) as part of a long range effort to

construct just such a light curve. Of additional interest

is the possibility that this star may also show evidence

of small amplitude changes similar to those discussed


The work is hampered primarily by the faintness of

LP101-15, (V =12.9, AV=0.75) and secondarily by its
flare characteristics. With a relatively faint (for

photoelectric observation) star such as this, one must use

the largest available telescopes as well as special observ-

ing techniques and equipment if optimum results are to be


Eggen and Sandage find no evidence of line doubling

or of any fluctuation in the radial velocity of LP101-15

from the spectra they obtained, which implies that the

companion may well be of considerably less mass than the

primary star. It would then be cooler and redder also,

perhaps contaminating the photometric data for photometric


Based upon their best data, Eggen and Sandage found

R-I=+1.32 in the Kron, Gascoigne and White (1957) RI

system, which gives Mv=+13m005 using the (Mv,R-I) relation

as determined by Eggen and Greenstein (1965). If one uses

Joy's determination (1947) of the H-R diagram in this

spectral range, the result is M =ll5lT for spectral

type dM4, and this is the value adopted by Eggen and

Sandage since they feel some contamination is occurring.

This yields a value for the photometric parallax of n=0.063

arc-seconds, or a distance of 15.9 parsecs.

More will be said about this system later, but it

is obvious that it exhibits a potential for yielding a

great deal of useful information with further study.

Chapter 3


One major aim of this project has been to design

and have constructed a complete photoelectric photometer

suitable for use in sky limited conditions such as are en-

countered when objects such as those described herein are

under study. However, a basic requirement from the begin-

ing was versatility, to prevent unnecessary duplication of

equipment. For instance, it should be possible to use the

instrument to study very bright stars as well, thereby

implying a range of approximately 15 magnitudes in object


During the design stage, as well as for the bulk of

the construction phase, though, it was obvious that avail-

able equipment would have to be used if observing was to be

done. For this reason, it was necessary to locate and

eliminate some persistent areas of trouble which had plagued

the apparatus previously, and finally to calibrate it for

standard UBV observations.

Existing Equipment

The original photoelectric photometer available for

use at the University of Florida's Rosemary Hill Observatory

on the 75 centimeter reflector was a dual-channel instrument


built by Astro-Mechanics, Inc., which has been previously

described in several publications (see,for example,

Bloomer (1973)). It was used unmodified after careful exam

nation and some minor adjustments made it possible to

successfully use dry ice for refrigerated operation.

Initially only UBV observations were carried out

while a careful analysis of the requirements of the problem

was made to determine which wavelength bands would divulge

the most information. As already explained, this led to

the decision to concentrate on a UBVRI system, when it

could be established.

Due to photomultiplier limitations and lack of

filters, at first only "R" observations were possible. "R"

will be used throughout to denote the pseudo nature of this

band as used at Rosemary Hill. To obtain "R", actually a

fairly close match to Johnson's R, an S20 photomultiplier

was used in conjunction with the dichroic beam splitter

for the necessary short wavelength cut-off. Work with this

system confirmed the feasibility of working very faint

objects, essentially down to the sky limit, with a well

designed system. It was then decided to proceed with the

design of the new faint star photometer.

New Equipment Required

The apparatus described above functions adequately

on objects of V less than approximately +8, but for fainter

sources, several problems become apparent. Primarily, the

sky brightness becomes a significant portion of the signal

received. As stated above, by V1ll, the signal to noise

ratio in "R" is nearly two, operating at dry ice temperature

with an S20 photocathode, while the sky contribution with

a 30 arc second diameter diaphragm averages approximately

25% of the total signal. This would be true for an object

of V-R near +1.0. Furthermore, considering the desired

application to red dwarfs with B-V=+1.5 and U-B=+1.2, it

is obvious that if V=+12-13, the sky brightness will dom-

inate in U and B unless the telescope used is larger than

90 cm aperture or the sky is very dark. At Rosemary Hill

Observatory on a dark, moonless night of average transpar-

ency, the ratio of the total signal in V to the sky contri-

bution in V reaches two around V=+13 with a 30 arc second

diaphragm and an Sl1 tube operating at -1000 volts. If

this object happens to have B-V=+1.5 and U-B=+1.00, then

U=+15.5, which typically results in a total signal to sky

signal ratio of perhaps 1.2, usually less. If the trans-

parency is poor, this can easily become essentially unity.

Incidentally, it must be pointed out here that most of

even this response is due to the red leak (see Chapter 8

for discussion) of the U filter, which is most significant

for very red stars.

The point to be made here is the magnitude of the

sky contribution and the important role it plays in the

possible accuracy of any observation. Any change in sky

brightness or transmission in these ranges appears

immediately in the output, and must be accounted for if

accurate results are desired.

Standard photometric technique calls for making

a set of observations of star plus sky in each filter, this

being followed by an equivalent set of readings on the sky

alone (normally obtained by driving the telescope off to

a nearby area judged free of stars). Each filter may be

used for one minute on the star and from 15 seconds to

one minute with sky alone, with the time between star plus

sky and sky alone being then at least two minutes, probably

three or more. Of course, these times are subject to a

great deal of variation depending on the brightness of the

star under study, but obviously a great deal of change is

possible over a time span of several minutes.

The end result desired then would be a reduction of

the elapsed time between star and sky observations, for

which several approaches are possible. First, one can sim-

ply take sky readings immediately after the star observation

in each color. This introduces the extra delay of driving

back and forth from sky to star, but is the only improvement

possible with standard equipment.

A second alternative would be to make some provision

to more quickly oscillate between source and sky, by instal-

ling some type of chopping system for example. This

approach will be discussed later in more detail in associ-

ation with work done at the Kitt Peak National Observatory.

Suffice it to say here that this method is capable of

contributing greatly to accuracy, especially when the

frequency of oscillation can be made high enough to give

essentially a constant check on sky brightness. There are

several problems with this approach. There is still time

used to observe the sky alone and the readings are not

truly simultaneous. In addition, somewhat different areas

of the filters and photocathode may be used for the

respective measurements if the correct optics are not


Third, truly simultaneous observations may be

obtained by using dual apertures, with various possible

beam splitters following each aperture to give a wider

spectral range covered, also simultaneously. This approach

of course requires the use of two or more detectors, which

must be as identical in response as possible if they are to

be used in the same wavelength region, or with the same

filters. Here it becomes possible to use as many channels

as may be desired, limited only by economics, the bright-

ness of the stars under study, and the weight capacity of

the telescope. Of course one must be prepared to do

extensive calibration of the various detectors, and a great

deal of care must be given to electronics design for data

collection. Furthermore, the pairs of detectors must

receive the same driving voltage (perhaps by driving two

in parallel from the same power supply) to prevent anomalous

color response changes with drifting voltage. In addition,

it becomes mandatory to use some type of tube refrigeration

to insure equal and constant photocathode temperatures.

For all its difficulties, the latter method provides

the only way to obtain simultaneous measurements of star

and sky, and for truly sky limited observations this has

been considered of overwhelming importance. It was there-

fore decided to design a new photometer for observation of

very faint stars, but one which could also be used for

bright star work if desired.

First, to provide a good base upon which to build,

an offset guider was designed to carry the photometer to

come, as well as many other pieces of equipment. A rapid-

offset mechanism was incorporated in the design to make

possible accurate and reproducible motion from a star to

a sky region for single channel or similar operation.

Great care was exercised in the design of this apparatus

(as with all of the new equipment) to insure freedom

from light leaks as much as possible.

The photometer design was finalized with two pairs

of channels possible for simultaneous star-sky work, plus

a separate access to single apertures for bright star

single-channel observations. Each access port would be

provided with a six hole filter wheel for spectral isolation,

with corresponding pairs being driven simultaneously

and remote control possible.

It was also considered necessary to design a dry

ice cold box suitable for use in the high humidity condi-

tions so often encountered in Florida. Compactness and

minimum weight were of prime importance here, as was easy

access to the photomultiplier itself.

Further details, as well as schematic drawings of

the individual pieces of equipment, are provided in the

chapters to follow.

Chapter 4


As stated earlier, this system was intended to be

complete in itself for almost any general photoelectric

(stellar) application. This first required an analysis

of the requirements of the various possible observational

programs which might be undertaken using the system and

finally a transfer of these requirements into mechanical


The first major component to be designed and con-

structed was an offset guider, one which incorporates

a few highly desirable refinements and capabilities.

The Offset Guider

This device serves several purposes. Primarily,

of course, it provides a wide field acquisition utilizing

the full aperture of the primary telescope to which it

is attached. This full aperture is absolutely necessary

for locating very faint stars and centering them properly

for observation.

Furthermore, if an object is too faint to be

seen even with the full aperture of the telescope, it can

still be observed if it is possible to locate its position

correctly. The star can then be positioned in the diaphragm


and observed for an extended time. The only means of

accomplishing this is by working from the positions of

visible objects. Using the equatorial coordinates of the

desired source in combination with those of nearby stars,

the position of the invisible star can be located in the

field of view provided the orientation of the system is

known with respect to the north-south direction. Using

this orientation angle, a set of transformation equations

can be set up to give rectangular coordinates for the

desired location, referred to the positions of the brighter

stars. The only requirement then is an ability to precisely

locate the center of the field of view with respect to the

point in the field seen bythe detector.

To accomplish this location, the field viewing

eyepiece is mounted on a movable stage with precise

readouts of its position in rectangular coordinates. To

provide reasonable speed in moving the eyepiece as well

as acceptable precision of location, one-half inch

diameter, 20 threads per inch screws were adopted. These

were made long enough to allow a total travel in both

axes of 2 1/4 inches (+1 1/8 inches with respect to the

center of the field). This corresponds to approximately

15 minutes of arc total travel. The positions are read

out on identical dial indicators to a precision of 0.001

inch or approximately 0.4 arc-seconds. Allowing for

some mechanical error, a reasonable estimate on the possible

accuracy of any position would be 1 arc-second.

The mirror allowing access to the optical path is

1/2 inch thick floated plate glass which tested reasonably

flat without polishing. It was simply aluminized and used

after an elongated 5/8 inch diameter hole was bored at

45 degrees to the surface to allow the beam to pass through.

The hole had to be elongated since it was intended to be

used with dual as well as single apertures. No vignetting

occurs for dual apertures up to 120 arc-seconds diameter.

Position control for the mirror is by means of a push-pull

rod with a two position ball stop. Optical alignment

is also possible by means of two spring-loaded screw

adjustments which are accessible through a port in the

bottom of the guider.

It should be pointed out here that light baffling

is provided throughout the guider. The movable stage

carries a two-piece metal baffle which completely covers

the port into the optical path. Also, a thin-wall,

elliptical, metal cylinder fits into the hole in the mirror

described above, and extends back to the rear wall of the

guider where accessories attach.

One of the more important features incorporated

in the final design is the rapid offset facility. The

mounting plate to which accessories are fastened is

supported on two cylindrical stainless steel shafts

inside the guider by means of four precision ball bushings.

These bushings are in turn attached to a plate which is

free from the body of the guider (the shafts alone are

fixed to the back wall of the guider). This interior

plate is controlled and located by a 1/2 inch diameter,

20 threads per inch screw, driven by a knob outside the

case. A spring-loaded ball rides on a disk (attached to

the screw shaft) with four equally spaced stops to give

reference positions for precise re-location (to within

one second of arc). A millimeter scale is provided for a

rough position readout, since more accuracy is not needed

at this point. A dial indicator is available if the

increased accuracy should be required.

The guider is intended for multiple application.

It has been used for some time with the Astro-Mechanics

Dual-Channel Photometer, as well as a special three-

channel system used for flare star observations. It was

primarily intended to carry the full four-channel system

(described later in this chapter), however and was designed

with that in mind. For four-channel work, no sky offset

will be needed, so to provide extra strength and stability

it is possible to bolt the accessory mounting plate

directly to the back wall of the case of the guider. If

this is done, the control knob should be removed to prevent

damage to the interior drive mechanism.

Finally, the focal plane can be located anywhere

from flush with the surface of the mounting plate out

to two inches beyond the surface of the plate. Beyond this

point, there is vignetting at the hole in the field viewing


Comparison Star Monitoring

Although this application of the offset guider has

not been pursued, the possibility exists for making

provision for monitoring the brightness of a comparison

star simultaneously with observations of a program object.

Sky brightness would also be simultaneously determined by

use of the full four-channel system to be described later.

As stated above, the wide field eyepiece is mounted

on a movable stage whose position is measured by twin

dial indicators. The prerequisite for monitoring a

comparison star is then that the star be situated close

enough to the variable to be within the range of the offset

motions. This places a rather severe constraint upon

selection of a comparison object, but in most cases a

suitable star can be found within the seven minutes of

arc radius restriction. Of course this probably will

not be true for bright stars.

During on observing run, then, the movable stage

would be set up to carry the detector so that it could

be moved at will. To accomplish this, a side-window

photomultiplier of small physical overall size (such as

an RCA 1P21) would be most easily adaptable. A compact

tube housing would also have to be designed to interchange

with the viewing eyepiece on the stage. Since little

weight can be carried by the stage, dry ice refrigeration

would be impossible unless extra support could be provided.

Possibly some sort of dewar could be built to allow

refrigeration with liquid nitrogen if it appeared necessary.

One or two filters could be used without excessive diffi-

culty to provide some flexibility, though caution should

be exercised to prevent unnecessary complexity. It must

be borne in mind that the primary purpose of monitoring

a comparison star in this way would be to provide a

continuous check on atmospheric transparency. For very

accurate information on color changes and the like, it is

still necessary to observe both variable and comparison

stars with the same photomultiplier.

Allowing for the above mentioned difficulty, having

a constant monitor for transparency would permit very good

time resolution. That fact alone would make the total

system much more flexible in various observing programs.

For instance, such good time resolution would be most

valuable in the study of very short period eclipsing

binaries or other variable stars which show rapid fluctu-

ations in brightness.

Finally, since the instrumentation would be very

simple and relatively inexpensive, the rewards would

greatly outweigh the effort necessary to develop the


Multi-Channel Photometer Head

Figure 4.1 shows the interior arrangement incorpo-

rated in the final design. It should be emphasized from the



*) I

04- 0

.,j l

41 C bj
* I -H
*. 0.0 0

x ) cl r-
0 a

< Bm - H -0

41 0


cn Iin 41c
4- .0

-I~~~44 F-'- Ia

0z u

4 nn


m 0-40

04 -0
bf~l 0

0 4-U)
r '00





first that ease of operation and versatility were important

considerations in planning, especially for this piece of


Wherever possible rotating devices were used

instead of push-pull types. Ease of access to (and inter-

changeability of) filter carriers was also important, as

was allowing for future automation of filter changing.

Actually, all important parts are either easily accessible,

by removal of no more than one side panel, or are themselves

completely removable from the chassis.

First consider the apertures. A rotating (manually

controlled) stainless steel disk carries up to ten small

brass disks in which the apertures have been machined.

These brass disks screw into the large stainless steel

carrier, thereby being easily replaced if different

size diaphragms are desired. Positioning is accomplished

by means of a spring loaded ball with positive stops.

Some alignment of the apertures with the optical axis is

possible since the ball-stop device can be shifted slightly

in position. See Table 4.1 for a list of aperture sizes.

Following the apertures is the diaphragm viewing

apparatus. This provides a sufficiently wide field to show

both openings in the dual diaphragms though the holes are

seen to be close to the edge of the field. Two identical

achromats are used as transfer lenses, being mounted within

a push-pull brass cylinder (which rides in a carefully

fitted stainless steel sleeve). When this brass cylinder

Table 4.1

Aperture Sizes

Aperture No! Dia.(mm) Drill No? Field Dia?(sec. of arc)

1 0.610 73 10.0
2 0.838 66 13.8
3 1.191 56 19.5
4 1.702 51 27.9
5 2.438 41 40.0
6 3.454 29 56.6
7 4.195 10 80.6

8 1.191 56 19.5
9 1.702 51 27.9
10 2.438 41 40.0

'Apertures 1 through 7 are dual apertures, with a center
to center spacing of 9.525 mm (0.375 inches). Numbers
8 through 10 are single apertures with the holes placed
on the optical axis. These are only original equipment
and may be subject to change on demand.
2This is the reference number indicating the actual
machinist's drill used to generate the aperture.
3Calculated assuming an image scale of 16.4 seconds of arc
per millimeter.

Filter Wheel
Assembly /
}<---Drive Shoft

Top cf Photometer

(Not to Scale-Schemotic)
Figure 4.2. Schematic drawing illustrating the proper in-
stallation orientation for filter wheels.

is pushed down to its stop, a mirror (fixed on the end

of the brass cylinder, but adjustable) is inserted into

the optical path to provide access to the apertures. In

its other rest position, pulled out against another stop,

the mirror is retrieved into the supporting sleeve. A

disk is attached below the mirror so that when the mirror

is in the "up" position, an effective light seal is provided

for the viewing assembly. Of course alignment is important

here, so a guide-pin and groove arrangement was provided to

insure against rotation and to give limit positions. One

should never try to rotate the brass cylinder since such

action would shear off the guide-pin. Nylon screws are

provided for two purposes: one serves to lock the viewing

eyepiece in position of crude focus (fine focus being

accomplished by rotation of the eyepiece), while the other

is a combination locking and pressure screw for the motion

of the main cylinder. At times, especially when it is

cold, the unequal thermal contraction of the stainless

steel and brass may cause too much slack so that the eye-

piece will not remain in the "up" position. The screw may

then be tightened to the correct pressure. Simple wear

on the contacting surfaces will necessitate some further

compensation. Probably the best arrangement would have the

thumbscrew driving against a short spring, capped with a

Teflon pad to bear against the sliding surface.

With the viewing assembly in the "up" position used

during an observation, one of two things may happen. First,

if the photometer is set up for single-channel operation,

the beam from the single aperture will pass on back to

transmission filters which define the light seen by the

photomultiplier. Finally, the light passes through the

Fabry lens (mounted at the front of the coldbox) and

on to the photocathode.

If the system is set up for multi-channel work with

dual apertures in place, the beams strike the faces of a

right angle aluminized prism situated behind the viewing

device. Each face is at 450 to the optical axis so the

two beams are diverted 900 in opposite directions. There-

after the equipment is symmetric in design and layout.

A variety of dichroic filters can be situated in each

beam to further re-direct and divide the beams by wavelength.

With these dichroic filters being inclined at 450 to the

optical axis, generally red light will be transmitted

without deflection, and blue will be reflected--both

transmission and reflection occurring with high efficiency.

Assuming identical dichroic elements, there are then two

pairs of beams available for study simultaneously. A

variety of dichroic filters will be available on the

carriers, as well as an aluminized mirror and an open


Filter wheels are installed in each beam, being

driven in pairs. At present, manual drive is provided for

these wheels through two identical Geneva mechanisms.

One complete rotation of a knob generates a 600 motion

at the wheels, and with the control knob properly located

in its rest position the Geneva will prevent further

wheel movement. Motor driven filter changing could be

accomplished simply by extending the manual control shafts

through to the exterior of the photometer case. Motors

could then be attached, preferably in such a way that they

could be easily and quickly disengaged if manual control

is to be preferred.

Users of the photometer should be aware that the

filter wheels have a single preferred orientation for re-

moval or installation (six orientations are possible).

Assuming the controls are properly set, a wheel can

only be inserted if the axle is rotated so that the leading

edge is as shown in Figure 4.2 on page 35.

After passing through the filters, the beams leave

the photometer head, passing through exit ports into the

individual photomultiplier housings. These ports may

be individually sealed if not all channels are in use.

Light baffling is also provided at these exits, since

they are especially susceptible to leakage.

Photomultiplier Chambers

For the applications intended with this equipment,

refrigeration of the detectors is absolutely necessary.

Previous experience showed that in the high humidity

conditions prevalent in Florida, serious problems are to

be expected from moisture condensing upon strategic

points--causing extremely high background signals. For

this reason, and since compact, light-weight cold boxes

were needed, a special effort was made in their design

to insure optimum performance.

First, for ease of removal and mounting, all the

boxes are to be supported by four thumbscrews. Location

is made simple by the light baffle arrangement. Inciden-

tally, all features (externally) of the boxes are identical,

so that they are all interchangeable physically.

Dark slides are provided with each one, with the

slide situated directly forward of the Fabry lens. Tests

show these slides to be sufficient against normal dome

lights with the photomultiplier at operating voltage,

but this practice should not be condoned except as abso-

lutely necessary. If lights are to be turned on, be

certain the slides are entirely closed. Only if they are

entirely shut can the slides protect the tubes against

excessive light levels.

The Fabry lenses are single element plano-convex

quartz optics, of 1 1/4 inch diameter and 2.1 inches

focal length at 5000 A. They are sealed in place with

thin Teflon washers, pressure being applied by a screw-in

mounting ring bearing against the convex surface. The

rear (plane) surface of the lens rests against a flat

supporting ledge. If the lens is ever removed, the seals

will probably be destroyed, so removal should be a last



Figure 4.3 illustrates the interior of the cold

box to a first approximation. Briefly, the aim was to

make the entire assembly such that it could be broken

down if necessary, while preserving its integrity when

in use. To provide the required seals at all joints,

rubber 0-rings were incorporated liberally. Where possible,

even complex parts were machined from solid stock to elim-

inate a joint.

The interior chambers are made of copper, turned

as thin as possible to retain the necessary strength. For

support of the inner chamber and access to the outside,

sections of stainless steel tubing were machined to

0.005 inch wall thickness except for approximately the

last 0.125 inch of each end of each piece of tubing.

This length was left full thickness of 0.025 inch wall.

This extra thickness gave added rigidity, and permitted

soldering the tubing to the copper inner chamber. The

thinwall section reduces heat transfer to a minimum,

though a material such as Inconel would have been better.

As can be seen on the drawing, the photomultiplier

may be removed simply by detaching the cover-plate on the

end of the box. The socket is mounted directly to this

cover-plate by means of nylon screws and Lucite spacers.

A heavy Teflon ring rests on the Lucite spacers and carries

two spring loaded supports for the socket itself. Since

the position of the photocathodes will be the same,

regardless of the photomultiplier in use, the Lucite

spacers will have to be changed if tubes of different

length are used. With all EMI tubes presently in use,

the entire length of the photomultiplier is within the

inner copper cylinder. Only the rear of the socket

protrudes into the stainless steel cylinder.

Teflon spacers are provided for support of the

front photocathodee) end of the photomultipliers. These

spacers can be easily modified to accept the addition of

magnetic lenses should they be desired to further enhance

the performance of the tubes.

As a convection baffle, a thin glass disk (0.040

inch) has been installed at the junction of the small

stainless steel tube and the front plate of the coolant

chamber. This glass was cut from a photographic plate

after having the emulsion removed. No seals are provided

for it, and location is accomplished simply by means of

two brass slip-rings. The glass is placed between the two


Four power resistors are connected in parallel

and spaced equally around the Fabry lens on the inside

surface of the front plate of the cold box. With appli-

cation of six volts, these resistors dissipate sufficient

heat to keep the lens frost-free. Probably less heat

would suffice under dry conditions. Heat transfer to

the cold inner chamber from these resistors is minimized by

the thin stainless tube, and by the maximum possible

thickness of insulation at that point. Incidentally, a

thermistor was put into thermal contact with the outer

end of the small stainless tube (connected to two of the

pins of the socket which supplies power to the heaters).

It should eventually be possible to use this thermistor

to control the voltage to the resistors, thereby regulating

the power input.

Finally, the center pin of the socket described

in the paragraph above is a ground terminal. From this

pin, a thin (to minimize heat loss) wire runs and is

soldered directly to the large outer copper shell of the

cold chamber. It is a further shield against interference

for the photomultiplier.

Santocell insulation is used in this design, and

appears to give good results. A full charge (approximately

six pounds) of well crushed solid CO2 will last for a

full twelve hours (if "topped off" after five or six hours)

of observing with no significant temperature change at the

photomultiplier. Under very warm conditions, this may

not be true, requiring the addition of more coolant.

In most cases, a temperature change shows up

immediately in the form of an increase in dark current.

However, with S-1 photomultipliers the sensitivity of the

photocathode appears to change even before the dark current.

Therefore, with the latter type of tube, it is recommended

that close watch be kept on the coolant. If the ice recedes

more than 1-1 1/2 inches from the front of the chamber, it

should be refilled immediately. This can be done very

quickly, and does not require a delay for temperature

stabilization since no real change will result from such

a minor addition.

In conclusion of this section, a word of warning

to anyone who may be dis-assembling one of these boxes:

use great care to avoid excessive force at all times and

watch to see that none of the small wires (such as the

ground wire to the copper chamber) are broken. While

it is true that dis-assembly is possible, it should not

be done unless really necessary due to the possibility

of damage. The thinwall stainless steel tubes are

particularly difficult to make, and replacing a damaged one

would be a very major operation. Properly handled, these

boxes should last indefinitely and perform well, as indeed

should the entire photometer.

Data Acquisition

At the present time, data are obtained by several

means at the Rosemary Hill Observatory. A dual channel

chart recorder is available for visual display continuously

and a digital system can be used for numeric output.

Presently, the digital equipment (mostly constructed here

at the Astronomy Department) gives a visual display of the

total counts per second, and records it on magnetic tape.

So, it is possible to hand record the display of the digital

unit, rely upon the tape recorder, or simply use the strip-

chart record of an observation.

Also recorded on the tape are the Universal Time

of the end of an integration and the settings of a rack of

manually set switches. Thus one may place an observer code

number, Julian Date, object code, or the like, on the tape

with each observation.

In the future, when full four-channel or even five-

channel work is contemplated, the magnetic tape recorder

will become essential. This is true unless a small com-

puter could be connected in-line to handle the massive vol-

ume of data which generally would be forthcoming. Indeed,

with the full system operational, a computer will be a

necessity, whether in-line or used with the tape output.

This is obviously true where high time-resolution work is

involved over an extended time period.

For computer handling of data, however, additional

information must be recorded. This includes filter identi-

fication, amplifier gain settings, and object identification

codes. Such items as sidereal time and local hour angle

are not necessary, being easily derivable from the UT plus

object coordinates. Also necessary would be the length of

an integration--unless counts per second are recorded origi-


Switches for indicating object codes, which would

be set manually, should of course be located where they

would be easily accessible to the observer while working at

the telescope. On the other hand, the filter identification

and amplifier gains should be handled by automatic encoders

of some type so that errors do not occur. There should

probably be a record of the dichroic filter selectors'

settings also, as well as at least one record of the

aperture used. The latter two could of course be manually

recorded, but there could perhaps be occasions when an

automatic recording would be preferable.

Clearly, the system as outlined in this chapter,

both existing and planned, holds great promise. It is

easy to visualize a completely automated observing program

in which little human error could be introduced, and

which would yield maximum results in the time available.

Chapter 5


This chapter describes in detail the techniques used

in making the observations to be presented in Chapter 7.

Since the facilities of three observatories were utilized

in the course of this investigation, the discussion will be

broken into units to illustrate the unique character of the

various instruments used and the manner in which available

instrumentation dictated the technique employed.

Kitt Peak National Observatory
Existing Equipment

The data obtained at the Kitt Peak National Observa-

tory, Tucson, Arizona, form a set unto themselves, and are

considered separately in the discussion of data analysis

and results. Since it was not feasible to have the pre-

viously described apparatus completed and operational prior

to the observing run, it was decided to utilize equipment

already operational belonging to the Observatory, which

further emphasized the desirability of considering these

data as a distinct unit.

Basically, the apparatus consisted of a modified

three-channel photoelectric photometer of standard Kitt

Peak design plus an offset guider and two end-on photomul-

tipliers mounted in dry ice cold boxes.

The offset guider was another Kitt Peak design, but

was not actually used as a guider. Its function in this

run was to act for field acquisition and centering, being

set up with a 25 mm focal length eyepiece throughout,

giving adequate field and magnification.

As emphasized earlier, the motivation for developing

the photometric system described in the previous chapters

was to allow the acquisition of sky and star plus sky

signals simultaneously so as to reduce to a minimum the

effect of changes in sky brightness with time. The ap-

proach followed with the Kitt Peak equipment was directed

toward this same point, but the method was different. The

standard three-channel photometer head was modified to

accept a chopping device forward of the apertures (two

nearly identical apertures were installed, oriented north-

south), which covers first one, then the other, alternating

between the two 600 times per minute. The chopper is

actually a disc with two concentric semi-circular slots of

different radii, 1800 out of phase, the disc being driven by

a 600 rpm synchronous motor. One available option (not

used for this work) was the ability to choose a different

slot arrangement, to give an observing time per aperture

ratio of three to one (the method used was one to one).

A gating circuit sends the pulses from star plus sky and

sky to two separate channels in a totalizer which then

prints out an accumulated signal for each channel plus the

integration time.

It must be realized that for any integration time

only half is spent on either star plus sky or sky alone.

So, the only time saved is that normally required to offset

to the sky manually, with the primary gain being that the

two readings have been taken very close together in time

(which removes a great deal of the difficulty in observing

relatively faint objects). It should also be noted that

this method also works fairly effectively for brighter

stars; however, since it is a pulse-counting system, a loss

in accuracy occurs with very bright objects with respect to

DC charge integration techniques due to a loss in linearity

of response.

Another limitation encountered was the fact that

this system was set up to work one channel (photomultiplier)

at a time, with each filter being manually selected. An

aluminized mirror was used to deflect the beam to the second

photomultiplier, a switching arrangement being activated

by this mechanism to connect the totalizer and printer

to the correct tube.

Despite the obvious difficulties with some of this

equipment with respect to the observations to be made, it

functioned well and reliably, and after an initial period

of adjustment by the observer to some of its peculiarities,

the system was put to very efficient use. In fact, for

observation of objects of intermediate brightness (for a

particular system) this design is probably superior to any

other (allowing for some modification), due primarily to

simplicity of construction and calibration.

The detectors employed were end-on photomultipliers,

an S11 tube being used in UBV measurements, and an S20

for R and I. Specifically, the S11 was an ITT FW-129,

serial number 127104, driven at -1700 volts DC, mounted

in Kitt Peak cold box number 29, with the S20 being an

ITT FW-130, serial number 096806 driven at -1800 volts DC,

mounted in cold box number 1. Both boxes were used with

dry ice each night, no alcohol or other medium being added.

For filters, standard sets belonging to the Observ-

atory were employed, mounted in two six-position push-pull

slides. Pertinent data with respect to these filters are

given in Table 5.1.

Four sets of dual diaphragms were available for this

system, each being an independent unit which had to be

removed and a new one installed in order to change aperture

size. In all sets the holes were spaced 2 mm center to

center, with the diameters being as follows: Set 1, 0.44 mm;

Set 2, 0.63 mm; Set 3, 0.88 mm; Set 4, 1.25 mm. For this

work, Set 3 was chosen, since the seeing occasionally would

deteriorate enough to make a smaller size unsafe. The

equivalent field diameter with the Number 2-90 cm telescope

was 14 arc-seconds.

In order to judge the relative areas of the aper-

tures, as well as variations in photocathode sensitivity

over the surface, a calibration procedure was adopted

(similar to the procedure outlined in the user's guide for


0 rn t-
V) 0D I
0 00 0
Lo 7 cr









oO- o o-C oO co
00 00 00C 0 03 O1

LO to c N N

O tl N-

N N Nzl -

> ~P5 = n


oC C

0 0 0 u
u u U1




the instrument) which also was a means of detecting faint,

perhaps unseen (visually) objects in the diaphragm intend-

ed to yield the effective sky brightness. This procedure

consisted simply of observing a source first in one aper-

ture in all colors, then immediately re-locating in the

other and repeating the process. For safety, this was

done with the filter order being both symmetric and anti-

symmetric between the apertures. Furthermore, it was done

at least on one occasion for each source observed to check

for background objects, but the aperture selected for

standard use as the object channel was the northern one

(as seen through the viewing eyepiece), or that one corre-

sponding to the output of channel one on the printout of

the data.

As an additional check on the relative areas of the

two apertures, as well as on the transmission variations of

the filters over their surfaces, etc., several sets of

integration were run during evening and morning twilight

using the sky as the source. The apertures were checked

visually for stars and the telescope was directed to

various areas as an additional precaution against back-

ground objects. The accumulated results have been used as

a correction factor applied to the sky signals during

data reduction for all observations.

The start of an integration was not synchronized

with the chopper, and it was highly probable that any two

sequences would not begin at the same orientation of the

chopping disc. That starting point is actually irrelevant

if it is assumed that the duration of a rotation of the

chopping disc was constant, and that the integration were

precisely equal in length. Since the chopping disc was

driven by a sychronous motor controlled by a variable

frequency source, with a toothed belt drive linkage, no

slippage could occur, so the first assumption is reasonable.

With respect to the lengths of the integration times--

information supplied indicated that the timer had not

been calibrated by Kitt Peak personnel. Due to the sophis-

tication of the commercially manufactured unit, it was felt

such a calibration should be unnecessary. Furthermore,

the necessity of utilizing the unit each night precluded

the possibility of taking it to the headquarters for such

a check to be carried out. The times were set in micro-

seconds, and hopefully were correct and highly repeatable.

No systematic error has appeared which would indicate a

flaw here, but it is possible that non-repeatability

could account for some scatter in the data.

Unforeseen complications in equipment scheduling

forced changing from the sky-chopper to a standard three-

channel photometer set up for two-channel operation using

the same photomultipliers and filters for three nights.

The only change in observing technique was the necessity of

driving off to measure the sky brightness--this being done

always in an easterly direction from the object. With

this equipment, the procedure was to make one fifteen

second integration per color for objects brighter than

about V=9, and two fifteen second counts for fainter


For all measurements involving the star-sky chopper,

the filter order was BVU-RI for all objects other than

LP101-15. For the latter, it was VBR, the signals in

U and I being too small for the purposes of the investi-

gation. However, U and I magnitudes were determined for

completeness, and to provide a check on the U-B and R-I

data given by Eggen and Sandage (1966).

With the standard photometer the order was again

VBU-RI or VBR for an observation of a source, with the

reverse order being used for sky alone, immediately

following. Since this was also a pulse counting system,

no gain changes were involved and all of the sky readings

could be intercompared to look for changes over time

scales of several minutes. Any more rapid variations

necessarily went unnoticed.

The aperture chosen for use with the standard unit

was the closest match possible to that set used with the

chopper. According to the user's guide it had a diameter

of 0.87 mm or 13.83 arc-seconds assuming a scale of 15.9

arc-seconds per millimeter.

Briefly then, in conclusion of this section, the

observations made at the Kitt Peak National Observatory

covered a sixteen night interval from 21 June through 6

July, 1973. From 21 June through 30 June, the star-sky

chopper was employed and again from 4 July through 6 July.

The standard photometer was used from 1 July through 3 July.

The photomultipliers and mirrors used were unchanged, only

mechanical parts being varied with the change as well as

the data recording system. For the standard set-up the

latter consisted of two pulse-counters whose output was

printed on digital paper tape along with the local sidereal

time, integration time, and identification (star, sky,

standard, or program object). The integration times for

the latter system were chosen so as to give counts which

could be directly compared with star plus sky counts from

the chopping system to allow rapid checks on night to night

transparency variations, which were found to be rather

severe. See Chapter 7 for the actual results in tabular

form and for discussion see Chapter 8.

RI Observations from
Lowell Observatory

In addition to the observing run conducted at Kitt

Peak, time was granted at the Lowell Observatory for

obtaining R and I observations of BY Draconis, GT Pegasus,

and FF Andromeda. For this work, the new photometer was

employed, though not in its most complete form since a

major construction effort was required even to bring it to

a state of partial completion.

The system used at Lowell employed the completed

offset guider described earlier. The photometer head was

set up for single-channel operation and manual filter

selection, and one prototype cold-box for dry ice refriger-

ation was used. Three photomultipliers were available in

case of failure, but the S1 photocathode EMI 9684B (serial

number 5490) was intended to be the detector.

A set of filters for UBV work were available in the

filter wheel in addition to the R and I filters. For the

R and Ibands, interference filters designated BP-R (R band)

and BP-I (I band) were obtained from Infrared Industries,

Waltham, Massachussetts, while the UBV set were standard

Corning glass filters commonly used for this application.

The transmission curves of the R and I filters are

given in Figure 5.1 with the relative response curve

(obtained from the published quantum efficiency versus

wavelength data for the tube type) superimposed. The

tranmsission curves are plotted in units of transmitted

intensity with the incident intensity assumed to be unity.

Figure 5.2 shows the transmission curves convoluted with

the response curve of the photomultiplier, and permits the

determination of ineffective central wavelength and

bandpass for each filter in addition to defining the

system more fully.

The observing run covered the nights from 18 Septem-

ber through 30 September, 1973, and was quite successful.

Equipment difficulties prevented observing on the first

night, and greatly hampered efforts the second. UBVRI

observations were then carried out for two nights

It I .50

IS 1-I ---- I .0
5000 8000 11000

Figure 5.1. Plot of I (X) versus wavelength in Angstrom
units. The response curve for an S1 photocathode is super-
imposed, normalized to one full-scale.



- C)













0 o




consecutively using an S20 EMI 9558B until the Sl tube

could be used. After the S1 was put into operation, only

one night was lost to clouds, data being taken for the

duration of clear skies. Since a considerable amount of

UBV data is already in existence for the objects covered in

this run, it was felt more worthwhile to concentrate on

R and I observations. In the following discussion, only

the S1 tube will be considered.

The photomultiplier output was sent to a standard

Rosemary Hill Observatory DC amplifier, then put through a

voltage to frequency converter kindly supplied by Lowell

Observatory, as were the counter and the digital printer

for recording the data. A timer was built into the Hewlett

Packard counter for timing integration. This timer had

a maximum of 10 seconds possible, with the next range

being 1 second (each successive step being down by 1/10),

so a standard 10 second count was adopted for all observa-


Since scintillation and image size fluctuations can

adversely affect the accuracy of an observation, it was

sometimes necessary to increase the number of integration

per filter. This effectively reduced the error associated

with the mean taken over the series of counts. In general,

two ten-second counts were made for standards or program

stars brighter than R-+9, followed immediately by one ten-

second count on the sky. Sky offset was accomplished by

means of the facility built into the offset guider (see

Chapter 4). During RI work, then, one data point for

one filter would consist of at least two ten-second inte-

grations on object plus sky and one ten-second reading

on sky alone. For fainter objects, or if seeing was poor

(worse than 5 arc-second diameter images), the number of

integration sometimes ran to six for the object and four

for the sky.

With the 79 cm telescope used with the S1 photo-

multiplier, such a procedure yielded excellent results,

especially since the sky was generally very transparent

and seeing better than 3 to 4 arc-seconds (though it did

occasionally deteriorate considerably). Though admittedly

six is not a statistically meaningful sample, the standard

error obtained for an object of R=+ll and R-I=+0.5 was

typically 002 for a single observation and nearly 0.01

or better for the mean. Many times errors were consider-

ably less than this. For this range in brightness, the

ratio of object + sky/sky brightness using the smallest

available single aperture (19.54 arc-seconds) was generally

around 3 in R, and approximately 2 in I. The signal to

noise ratio was usually better than 6 or 8, with the noise

being essentially all contributed by the sky.

Dark currents were totally negligible after allowing

the photomultiplier to stabilize at dry ice temperature

for one hour. One half pint of isopropyl alcohol was

added to the dry ice to give better thermal contact with the

tube chamber and insure uniform temperatures in the

photomultiplier. With the ice charge being renewed at

midnight, no temperature change was evident over a twelve

hour night of observing. The dark current held steady at

very nearly zero on a gain setting of 12.5/0.0 with very

little spurious noise being contributed. This was for

an operating voltage of -1500 VDC (chosen to minimize

noise and give results close to standard R and I magni-


At the end of a night, the voltage was turned

down to -500 VDC and left on at that level continuously

until being turned up again prior to being cooled at the

beginning of the night. There was almost always a remnant

of ice left in the coolant chamber in the evening unless

the temperature in the dome became inordinately high

during the day. With this procedure, the tube functioned

flawlessly and stabilized considerably in its response

after a few nights of use.

No mechanical trouble was experienced and the

observing run proceeded successfully. The results are

presented in Chapter 7 and discussed in Chapter 8, along

with data from the Kitt Peak and Rosemary Hill observa-


Rosemary Hill Observatory

Only the offset guider has been used at Rosemary

Hill, so little needs to be said above and beyond the

contents of Chapter 3. Prior to completion of the guider,

sky readings were obtained by driving the telescope itself

off to star-free regions, then back to the object or off

to another star. This was sufficiently time consuming

that it precluded making sky readings immediately following

a measurement on a source, especially in multi-color

work. It was then most efficient to make all the different

color observations on an object before going off to make

the sky readings. After the guider was installed, pro-

cedures similar to that used at Lowell were adopted for

sky brightness measurements.

The Astro-Mechanics dual-channel photometer was

used throughout, UBV observations being made with an

Sll EMI 6256S driven at -1000 VDC, and "R" observations

being done simultaneously with an S20 EMI 9558B driven

at -900 VDC. The "R" capability was discussed in Chapter

3; the UBV filters were a standard set belonging to

Rosemary Hill Observatory.

Data recording has been discussed in Chapter 4

so it will suffice here to say that after it was available,

the digital system was relied upon for primary recording.

The dual channel chart recorder then served as a backup

unit and as a constant check on sky conditions as well

as a time check.

The number of integration carried out for a

particular color or source was varied considerably

depending upon the immediate conditions and the brightness

of the subject. Occasionally, as a statistical check,

long runs of short (two to five seconds) integration

were performed, but primarily the method was to make

two or more 10-20 second counts in each color on sources,

with at least half that number being devoted to sky

readings subsequently.

If anything out of the ordinary was seen to be

occurring (such as a flare), the system could be quickly

switched to automatic continuous recording with short

integration for maximum time resolution. For most work,

though, it was either manually or automatically triggered

for the desired number of scans.

When it was available, a Cerenkov source was used

as a check on the response of the Sll photomultiplier.

As a general rule, the S11 tube was allowed to see this

source while otherwise idle during changes between program

objects to provide a reference level assumed to be of

constant intensity. Since the source had to be mounted in

a filter wheel in order to be close enough to be seen

by the photomultiplier, it only served as a direct refer-

ence for one channel. To a degree, this could be extrap-

olated to cover possible large scale abnormalities in

either channel.

Regrettably, due to a shortage of observing time

and the necessary long time spans between nights available,

the data obtained at Rosemary Hill must be used carefully.

The data serve mainly as a check for phase shift from

season to season, but some of the material is suitable for

independent use. For instance, one major flare was

detected on GT Peg, and will be presented in Chapters 7

and 8.

After the faint-star photometer has been fully

completed and put into operation, much more potential for

research will be available in this brightness range. In

particular, the data acquisition rate should be at least

doubled, with an accompanying increase in the accuracy of

a measurement. This is of particularly great significance

in the study of such small scale changes as are being

watched for in this project.

Chapter 6


Three methods of obtaining output were used in the

course of this project: DC amplifier output recorded di-

rectly on strip chart; DC voltage-frequency conversion for

digital output; pulse counting with printed digital tape


Each of these methods required different handling

techniques since the format was varied in each case and the

nature of the output was different.

It should be stated here that after considerable

thought, the most advantageous method for handling the Kitt

Peak and Lowell data (both were on printed paper tape)

seemed to be by means of the Hewlett Packard 9810A table top

computer belonging to the Astronomy Department at the

University of Florida. This machine is quite adequate for

handling relatively limited amounts of properly formatted

data. Hand entry of data is required for the 9810A, but

it would have been necessary in any case with these outputs,

since the material needed to be on magnetic tape or

computer cards for more automated large computer reduction.

Most data from Rosemary Hill Observatory were also handled

on the 9810A since the full facilities of the present

digital system were not always available and consistency in


the reduction processes was considered desirable.

With the strip chart output, it was necessary to

read times and deflections from the chart, with reduction

to follow. Conventional methods (see Hardie (1962) for

example) were used in this process, so nothing further will

be said with respect to this task.

However, after digital output became available, and

in particular after it became possible to make long series

of short integration, it was obvious that a more auto-

mated reduction process would be far more worthwhile and

less time consuming. When it becomes possible to record

such information as gain settings and filter identification

code, fully computerized reduction will be a realistic

solution. Even at the present time, for some purposes such

an approach would have advantages in speed and accuracy.

Following is a description of the adopted reduction

methods as finally decided upon for the extent of the

current work.

Pulse Counting

As discussed in Chapter 5, pulse counting devices

are utilized at Kitt Peak, especially for faint star work.

For the work described herein, this pulse counting equipment

was used exclusively. With pulse counters no gain changes

of amplifiers or scale changes are involved, so there is

only the linear response of the photomultiplier over the

entire range of sensitivity to be accounted for.

Incidentally, the primary difficulty with pulse counting is

the rapidity with which pulses occur for bright objects,

causing a loss in linearity due to insufficient time

resolution in the pulse amplifiers and counters. Pulse

counting therefore cannot in general be used effectively

for very bright stars. However, for the program stars

and standards used at Kitt Peak, this problem was avoided

by suitable choice of stars for calibration.

Over the entire range, then, the output consisted

(for the star-sky chopper) of three printed lines: one

line for sky counts, one for star plus sky counts, and a

third showing the integration time used. Identification

numbers for objects, filter identification, and Universal

Time (to the nearest second for the beginning of the first

integration for a filter) were hand recorded on the printed

tape for reduction purposes.

The format of the output from the standard photom-

eter has already been described in Chapter 5.

Input for initial reduction then consisted of sky

count, star plus sky count, filter identification, and

the UT for the midpoint of the integration. Object code

numbers, integration times, and correction factors for

the apparent relative sizes of the apertures, etc., were

entered into storage as constants, as were coordinates

for the objects.

The output consisted of object and filter codes,

plus the instrumental magnitude, the local hour angle, air

mass, Julian Date, and phase of the observation (if a

periodic variable).

Reduction equations in the program were set up so

that the counts were referred to the maximum count for

the brightest star observed. A built in zero-point

correction gave the instrumental "V" magnitude of this

brightest object a value close to its standard V magnitude

on the Johnson-Morgan system. The form of this equation

was simply

mag = -2.51og10{(N,+s-Ns)/(N max'at)}. (6-1)

Here N += number of counts from the star plus sky channel,

N = number of counts from the sky, N = reference level
s max
count adjusted for zero point, a= correction factor for

the apertures, and t= integration time in seconds.

Air masses were computed from the UT of an obser-

vation combined with the right ascension and declination

of a source processedd to the year of the observation).

The standard polynomial approximation for air mass was used

X = sec z -0.0018167(sec z -1)-0.002875(sec z-1)2

-0.0008083(sec z-1)(6-2)

Here sec z= secant of the zenith distance for observation.

This output was then further analyzed for computa-

tion of transformation coefficients into the standard UBV

system of Johnson and Morgan (1953) and the RI system of

Johnson et al. (1966). Actually, the R and I observations

were only psuedo R and I since an incorrect photomultiplier

(an S20 photocathode) was installed instead of the Sl

normally used, and the filters were not the same ones

used in the original work.

For the UBV data, the following equations are basic:

K = k+k'Co (6-3)

K = k +k'Co (6-4)
c c c

m= m-(k+k'C)X (6-5)

C= C-(k +k'Co)X (6-6)

Here C= true instrumental color outside the atmosphere,

m= raw, uncorrected instrumental magnitude as observed,

C= observed instrumental color, X= air mass of the obser-

vation, k= first order term in magnitude extinction,

k'= second order term for magnitudes, k and kc being
c c
the corresponding color terms. Specifically, if 1,...,j

designates a particular color or filter, C1,2=ml-m2, etc.

For transformation, we need m? and Co (for all

combinations desired), so k, k', kc, and k' must be ob-
trained. Several methods are available for this, so brief

description will be given of the process used. First, since

observations of standards were made for several nights, it

seemed desirable to combine them if possible for a more

accurate determination of the terms involved. Let i be a

particular point in one night, and let c identify a partic-

ular color, with n being a number to indicate a particular

night. Then Zn = N = the total number of nights. Rear-

ranging (6-6) yields (where Kc= k +k'C>O0)
c c c

C = C+K X (6-7)

which is in a form for a conditioned least squares approach

Forming the normal equations yields, after manipulation,

SZ(C)2= E(C+Kn X -C ) = 0. (6-8)
C i ni nc ni ni


Z ZE SC)2 = Z E(X .C+K X .-X C .) = 0. (6-9)
K nc n ni nnc n ni cni ini)

There are N equations of the form (6-9). After some manip-

ulation, one obtains for a particular night,

zx *C i-X .Co
Sni ni ni
K (6-10)
nc E2


zX .C .-EX .CO'
z Co= EZ C .-Z E-C X (6-11)
ni 2 ni
n i n i n ZX i
1 i

or, where E Z= total number of observations over all
n i
ZX -.C .X .
i ni ni ni

E EC .-E ZX2

i ni
ni n ni (6-12)

z E -E EX2
n i n ni

These values of Co are then to be used in (6-10) to

solve for the nightly values of Knc. It remains to solve

for values of m and Knm. Following the above pattern for

the C an Knc one finally obtains

EX -m .-zX .mo
ni ni n ni
K nm z (6-13)
nm 2

XX Znm .X .
ni i ni ni
1 l
ni 1 1

X (X )2-
mO= n i nl n (6-14)

Z Z -Z X2
n i n ni

Using equations (6-10, 12, 13, and 14), the best

possible overall values for K and K are obtained, as

well as for m and C. This is obviously only true for

those objects which were subject to frequent observation,

and other approaches must be followed for other, less well

observed program stars.

Problems may well arise in the application of equa-

tions (6-13) and (6-14) in cases where changes occur in

the sensitivity of the instrumental system with resultant

changes of the zero point. Hopefully, over a period of

a few nights this will not occur, but it must be watched

for carefully. Generally, the color ((b-v), (u-b), etc.)

sensitivity is quite stable, even over several months, so

the concern is not so great for (6-10) and (6-12).

Working from the values of K and K obtained

from the exhaustive procedure just outlined, and by using

the nightly observations employed in the computations,

one can solve for the first and second order terms in the

equations for extinction, (6-3) and (6-4). For instance,

if the natural system yields values for u, b, and v magni-

tudes, the desired values would probably be vO, (b-v)O, and


To obtain k _, the best method requires many

observations of a pair, or pairs, of stars of very dif-

ferent colors, each observation being made at nearly

equal values of X. Values of AC=(b-v)1-(b-v)2 (where

l=star one, 2=star two) may then be plotted versus X.

ACo is the intercept and k' ACo is the slope of the best
fitting straight line. Combining kv with the best

values of Knc then allows good evaluation of kb-v, the

first order term. The range in X needs to be 1-2 air


Johnson (1953) defined k' b=0, so the extinction

in U-B simply requires solving for the best fitting

straight line on a plot of (u-b) versus X. This yields

K ub=k u-=slope, and (u-b)o=intercept.

For k the same observations used for kbv are

sufficient. Values of Am =v-v2 should be plotted against

AC _X. From this, k' is the slope and (v -V )0 the
b-v v 1 2
intercept. Combined with Knv the value of k is readily


The above methods are directly applicable to RI

observations. Johnson (1965) has shown that results

obtained in this manner are the most accurate, though the

zero air-mass estimate of intensity is still generally

15 to 30 percent too low. The second order terms

(k' and k' _) are generally indeterminate, and the trend
T T-1
is toward lower values of extinction at these longer


With completion of these solutions, all the terms

are available for solving for the zero air-mass magnitudes

and colors for any object observed on a particular night.

Finally, it remains to make the transformation into

standard systems. If the instrumental and standard systems

are closely matched in filter and detector characteristics,

these are usually linear transformations of the following


B-V= Al+A2(b-v)O (6-15)

U-B= A3+A4(u-b) (6-16)

V-v= AS+A6(b-v) (6-17)

R-r= A7+A (v-r)0 (6-18)

R-I= A9+A0 (r-i)o. (6-19)

If the transformations are truly linear in char-

acter, one can apply a linear least squares technique to

solve for the values of Al, AZ,..., A10. This requires

formulating the above equations as follows:

A (B-V)+A'=(b-v)o (6-20)

A'(U-B)+A'=(u-b)O (6-21)

A (V-v)+A'=(b-v)o (6-22)

A'(R-r)+A'=(v-r)o (6-23)
A{Q(R-I)+Ag=(r-i) (6-24)

where the primed A terms are obvious functions of A

unprimed. In this form, doing a least squares fit to

plots of (V-v) versus (b-v)O, (U-B) versus (u-b)O, (B-V)

versus (b-v)o, (R-r) versus (v-r)O, and (R-I) versus

(r-i)o yields values of A>, A A6, A>, and A!0 from the

corresponding slopes, as well as values of Al, A', A5, A',

and AQ from the respective intercepts. It is then a simple

matter to obtain values of Al, A2,..., A10, which define

the transformation. As a result, one is in position to

specify values in standard terms rather than instrumental


In this manner, the pulse counting system was

calibrated to the UBV system quite well, but the R and I

fit was not so good, due of course to the lack of a proper

photomultiplier. Using the coefficients so obtained, the

observations of LP101-15 and BY Dra were placed on the

standard systems, and are presented in Chapter 7.


Since the format of the digital system output was

somewhat different from that used in pulse counting, the

reduction equations to obtain instrumental magnitudes were

correspondingly different. Additional changes were

brought about by allowing computation of standard errors

where several data points were taken in succession.

First, since DC amplifiers were used, gain settings

enter into the computation. The general equation to give

the instrumental magnitude was of the following form:

mag= S-2.51og10 i+s s (6-25)
a* t J

where S= amplifier gain (in terms of magnitudes), a= crude

zero point correction, and t= integration time. N,+s and

N are defined in (6-1), only here they are counts as

given by the v-f converter, counter set-up.

If the value of t is always the same, it can

be incorporated into a constant term, say b. Also, if

several points are to be combined into a mean magnitude,

let x = mean value of a set of (N,+s-Ns)/t values, so that

x (6-26)
mag = S-2.51og10 -. (6-26)

This is the form used to reduce the data from Lowell, since

all integration were of 10 seconds' duration and multiple

points were taken with each observation. If x1=(N,+s-Ns)I,

then the standard error for a single value of x, Sx, is

obtained from (where there are n observations)

2 2
(Sx) 2 Ex2 (x) 2 (6-27)
n-1 n(n-1)

and the standard error for the mean, x, is

Sx = Sx//n (6-28)

Equations (6-26, 27, 28) were utilized in the

reduction program for the Lowell data while (6-25, 27, 28)

were used for Rosemary Hill output. In the case of the

Lowell material, the program input consisted of N*+s, Ns,

UT, and filter identification as well as the storage

location for the gain used. Output consisted of (for n>l)

instrumental mean magnitude, Sx, Sx, object code, filter

code, local hour angle, air mass, Julian Date, and phase

(if appropriate) for the point. The latter four values

were computed by the same sub-routines as were used with

the pulse counting data. For the Rosemary Hill data, the

only change required in this procedure was the necessity

of entering values for t which were not always identical,

even for N*+sand Ns values obtained successively.

Following this initial reduction, the instrumental

magnitudes and colors of program objects and standard

stars were handled in a manner identical to that described

in the previous section under the discussion of extinction

and transformation equations. The final results for the

Lowell observations as well as those from Rosemary Hill

are also presented in Chapter 7 in the appropriate sections.

Chapter 7


This chapter contains the data collected at the

various observatories discussed in the text. Following

the tables are several figures, one of which contains data

in graphical form (the remainder of the figures are finding


Table 7.1 contains results from Kitt Peak for the

star LP101-15 while Table 7.2 summarizes the data for

BY Draconis (also from Kitt Peak). It will be noted that

not all of the dates on which observations were made are

included. This is due to the fact that reduction showed

the missing material to be of poor quality as a result of

cloudiness. Since no strip chart was available for an

indication of sky conditions, these doubtful observations

were made in the hope that results could be extracted (as

indeed some were). This points out one major difficulty

with either digital or pulse counting methods if no visual

output similar to a strip chart is available for reference.

During reduction for LP101-15, phases were deter-

mined for periods of 0.63398 and 1.26796 days in an attempt

to discern between the two. Therefore, two columns are

provided under "Phase", the first referring to 0.63398 day

and the second to 1.26796 days. Since only BV and R


observations were carried out for this object, the only

tabulated material is B-V, V and R in addition to the

Julian Date of the V observation and the corresponding

phase (calculated relative to Julian Date 2441855.750000).

See Lowell Observatory Bulletin No. 140 for a

finding chart for LP101-15 (G225-67). The comparison star

is shown on the chart, with approximate 1973 coordinates

(as measured from the Palomar Sky Survey prints) of

a= 16 56194 and 6= +57?04056. B.S. 6237 was used as a

check star, being located at a= 16h746282 and 6= +56782109

in 1973. LP101-15 is at 16h56555, +57?20279 (1973).

Table 7.5 gives all data as determined for the

check and comparison stars for all program objects. It

should be noted that all data are standardized to the UBVRI

system of Johnson et al. (1966) unless otherwise noted.

If accepted values are available, they are given unless

results obtained differ by more than 0.02.

For BY Draconis, the comparison star was DM+5102408,

and the check star was DM+512410 as given by Chugainov in

1966. Phases were calculated relative to JD 2441855.894271

for periods of 3.836 days and 3.826 days for the Kitt Peak

work, and relative to JD 2441947.727431 for the Lowell data.

Figures 7.2 and 7.1 are finding charts for GT

Pegasus and FF Andromeda respectively. Comparison and

check stars for each are indicated on the charts,with the

1973 coordinates being as follows:

GT Pegasus: comparison a= 22h84297 6=+31750543

GT Pegasus: check a= 22.84553 6=+31758993

GT Pegasus 22h84297 +31761377

FF Andromeda: comparison a= 0.68504 6=+35750951

check 0h68170 +35060120

FF Andromeda 0'68866 +35?40116.

These coordinates are only approximate, also being derived

from measurements from the Palomar Sky Survey prints.

Phases for the latter two program objects were

calculated relative to JD 2441947.940278 (for GT Peg) and

JD 2441948.943056 (for FF And) in the Lowell data. A

period of 2.17 days was used for FF Andromeda, while

several values were tried for GT Pegasus. The previously

published period of 4.65 days for GT Pegasus did not fit

the observations at all, nor did the previously suggested

2.0 day period. The best fit for the current data appears

to be approximate and near 1.20 to 1.30 days. It is per-

haps equally possible that the fluctuations in brightness

are not periodic at all. More observations over a longer

time span should answer that question. The 2.17 day

period for FF Andromeda appears to be correct.

See Table 7.3 for the results from Lowell. Here

also, not all nights on which observations were made are

represented, though the reason here is related to equipment

For the first two nights, a different detector had to

be used, and the results were too erratic for any serious

use. They were therefore omitted from the summary given

here. The dates given above for determination of phases

occurred during these nights.

Figure 7.3 is a plot of AI/I versus UT for a

flare observed on GT Pegasus at Rosemary Hill Observatory.

It occurred on the night of October 12, 1972,with the first

observation beginning at 04h34m53s UT. The signal was

almost back to the quiescent level by the end of the run

at 05h05m45s UT. On that night, observations were being

made in four colors--UBV and "R". The "R" band was being

obtained as described in Chapter 3, and data were being

recorded on magnetic tape. Integrations near 3.3 seconds

in length were used throughout the flare, with the elapsed

time from the beginning of one integration to the beginning

of another being almost exactly four seconds. The length

of each scan was recorded to the nearest 0.001 second along

with the UT at the end of a scan and the total counts

accumulated. The breaks which are seen in the data are a

result of taking sky and comparison star readings. Other-

wise, the "R" data are continuous, with the UB and V obser-

vations being made alternately with the minimum possible

dead time between filter changes.

As discussed in Chapter 5, the time between observ-

ing runs at Rosemary Hill Observatory was in actuality

too long to permit a detailed analysis of periodicity of

the small scale variations seen in these stars. For this,

several nights (extending to several weeks, if possible)

in succession are necessary, such as with the Kitt Peak and

Lowell data. The reason for this is the phase shift

seen to occur from season to season, and the occasional

change in amplitude of the light curves. As can be seen

from the tables and plots of the material obtained at Kitt

Peak and Lowell Observatories, the amplitudes for BY Dra,

GT Peg and FF And were all quite small in 1973, and the

scatter rather substantial. After several attempts to

collate the Rosemary Hill Observatory data with these

runs failed, the only conclusion possible was that no

major discrepancies exist and no significant difference

in behavior was exhibited in the 1972 or 1973 observing

seasons. One further complicating factor was the number

of equipment changes made in that interval--which forced

the devotion of much time simply to calibration runs. As

a result of these changes, however, the system is now ca-

pable of quite efficient operation.

During the summer of 1973, a few points were ob-

tained on BY Dra and GT Peg. Observations were made on

BY Dra on 20, 25 July, and 23 August, with GT Peg being

picked up on the night of 23 August. Only two points per

night were possible for each star, so suffice it to say

that they fell within the scatter of the corresponding

data from Kitt Peak and Lowell.

Table 7.4 summarizes the results from Rosemary Hill

Observatory. For each date, all observations were combined

(since most were made within a time span of 3-4 hours or


During the two seasons of observation of LP101-15

at Rosemary Hill Observatory, no minima were detected, and

the difficulty of working the star with the equipment

initially available made the data exhibit a significantly

large scatter.

With respect to the data from Kitt Peak and Lowell,

the tabulated values represent means of from two to six

individual points. The corresponding standard errors

for the means were usually quite small (less than 0.01

magnitude or better for LP101-15 in V, and generally near

0.008 in R for BY Dra, GT Peg, and FF And). The standard

errors for the colors (U-B, B-V, V-R, or R-I) were

correspondingly low--though the R-I values obtained at

Kitt Peak were quite erratic due to the low sensitivity

of the S20 tube in the I band.


t r 0 MirH tl o t l rtN- Ln fl i>-
Q D om o --tcoc "O ontrm"'t 0M0) m it0
Z-i 'tnt)lo) Lo t o D 1--i oL 'fl 't 0 t Of) l rNN-Qc

w ON-M d'C'fol0 tfCN Nmt N00
0 0000 Lnnfl N---- 0o omr
( ltaMf CO 1 0 E--4 o 0 n sLof o
v o co flf ll -:t- 0I'D m -i-n o r-rt- o -o1 o

oo co o Lon -C o Nr-C--N-C-c-No cN-o c -N-C

C r_- r-.Zl-o co ro -,oo I-- L -- i-n -n -

rn -o Tt -zt zi zj- -zi- Ln -n It

00)KmM CllO O O-O OO 1 010

O I oo-t tftlorto-40-1--0 v 0N -r0ioo r-- t 1-1c-
-4 1- mK c)oMcr Coi o o o c oo oo oo mc) m

0' Z4-' N-N-Otr t OO Ll- rO]

O oo o cc o c I-- o r- ) Lo lo l l cc lo i o

4 000 toN-O N- CCOOn m 4m L/v
GO 0000 N-N-N-N-N- 0o 0 D O l000 00 L t/

H0000 00000 00000 00 000

t C

cN r- LOC co -r t O-I N -OrN- cN-t to N-
n Lo M N-no -0- O0Ln t- 0ooCr-c)O c No D N-4L/ Ko
to cDm0 mtno~r r o r c vO co or

. . . . . . . . .
+ Ntt t/C 0 COCO0 0 'CN

0- CO) cN- --1 N- or- / C o -It Or N- 0o cc

tO LOt/-to rA tr4n-%D N-N-' cD D N-tn c N -yriN

r- N -- N-N- r- -- r r- oo 0o co c m r- N-

A Ln n LN-
o oo oo
tO t tO

00 r- 0) t)
tIn 0 Cn In t-
n 03 00 Ln 0

SND r- o00 0


*t- I-- Lr I t-
o C- 0)

t-- rn- LIn 0c In

In Ic In oo

o00o n IO
0 In tIn 0= "
N- 00 "J 0O In
In 000000

Ln Ln Lo Ln Lo


0 0
In -iH o 0 c
i- oC ID T- in

r-1 10 0- 0 0c
r-4 1-1 t-I i-


tn 10 NJ "-t
. t o I-1 00 M

r- r- r- N-


Ion 0 o0 o
0 In 0 In In
n It .t In

00 -1 0 Itn 0

NJ 00 0 r- -i

r-H -1 t-H C

NJ 00 tIn n 0D
In C) r- o0 -H

r- In m o o


i-n 00- 0o

00 m Nt- 0o
Ln In In r-1 0

c) "J r- "

In 0 oo :t

0c oo Cn 0 In
00- 00- oo o
N-N -0000

ar00 0
..... o

r o LO t
n- 0 In rot

vi -It- Ln 10 ^
- NN r


00oo r- o oo

O (1 r-t

tO (M ^ 0
000 0 -
-- oo -rI -

NJ Zn 0)- n-
o0 'D t-- i- oo

n 0 C) 0 0
000 In 0

oo r- oo Lo

00 0 oo oo t--

000 o r- Lo
0 0 oo

00o oo o00 oo oo


..... OC

NO D r- 0 tn

NJ 000N0

0I n -HL


C00 NI- -- 0

m oo ocoo

0LN- 0- 0

n000 LO L

SIn r- -DO --
SIn 00 I-M

oo I- = In
0oo r- N- n-

oI In (n Cn C)

0* *1N-
00 i NJ 00
00 N- (0 N-
0 'it00 N

o0 10 a n
) 0N 00
U) 0) oao o)

LN o o0 cUa

On n rN -
T-I LO 0

- ) r)
N o U) N ro

00 ro rN

\D) U) ) Nt n0
rN U) tnoo n N

[T-I COu T0 T--
N atn r N LO
t0 No an to 0
'r N on to Uo
N to to U()

7) Ln Nn 0-


U) N0 NOo


`0I N Un
Lo i i oo
o U) Z- L)
Scb v)
010o 0
o c) a I-


Cni 0 o) 0n -

Sc N cU) U)

S0 0 0

Ul l Ul) an -l

to N U) to 0)t

10 %C 11 -

N '- U) to N
tan to 0 't

in to L n o cq

on oo 0

r- t-o U) NU

r-- CD C) an a

000U) ) -
). r'- c n -1- )
CD 00 n 0


0 N r- 0 U-)
- rl- 1 Co Lr

1- -1 l

000 o n an an
U) Ci C7) rN NU
00 CO00 00

to C N C0
cn -1 CU ) ,i

-) U1-1 N N
r- a" U) -l c

C:N a)n rU ) UI

CN an a- an q0

U) r- 0- U7) Un
0 0 00
,COi4 cno "

0 d c0 o o o


0) 10 U) L- r-N
N-- r U)n to j-
U) U) to cN N
Sto n U) N-


t-l U)r N 0

0 N"U cU U)
n an _q a10 zr
Co No Cn U) eN
U7) cN al o N0

N N N) N
N 00 Ln C4
-Woo iH 0
Ln 1 or-- 1-0 O

N N N to to


N ro t Ni -


U)Nl to ano U


00 l
00 0



<* r-l


U) C)

O t



0 00

- to

N-~ N


r N 0r rL oC V 0 \0oa 0 o1m OC)n OOt-1 Ln \C
C'- Lnooc mrH H1-1CoCo -1 r Ln D r ZT ) L 'tM
C) -I O- Ln r- tf l,-i c'-H i fi fn r- o: (m tnt tt mf 0) m t n
V) dlz )\ LO o 00) t- C0 C -) tn r U') 10 ID> r- CO -Acq T- i- tDn Ln 10
S Mtn toMto l) 'rZt Ln Ln n LnL n ccoc00

fL 00000 00000 00000 0000 Co00000

ol r- r-4 lfln n )O 1 -1rH dlf -- l in0 i -4 o *r- f LO C
SCOr Mr- C 'D nr- f mLn tfl C)ldl Qio

tom cc D -D tn i-1Zr e-T f tno T-- o CD i- i- 0) 0) t-- r-e me rmn
ao m t o tfr"- 0C:, ] -t LfM Vi fC' 3-4 10 ) -Ln L1 mL Om
0co 00a 0) 00Q Lc-ooD oooo cL--o oo 0 00cr)C cc

'T CI N C ':)- rNC-I'rJ'Ct' D 't^N')' C D) Cc-1iCC 00c3 oCNCN
Hrh- in (: tn tn 10 Ln cD CD tn Tl- A- r -H r 0 -1 r-
*H i O LnfLnfl '9 0OLCoLf) Ln 't LO~n d\rIt n Ln Lnm O c 0LO

0 +

C-) C-]OrH ir fO l ir -zt C) oc)Ln 000or-~-
t]Q In Ct I ) l001tM t >- Ln OC)LOLC Ir'D1CC"l100
1 -n r-4 oo tf ti l0Occ "j' i-i r- <0i C 1 i-1'o tf0D 00c L ZTL

C) 1--N]r---0oC) o t)ltninL nLn \0oC--r-r--oo oCO S>0000 0M0 000000 0000 00000
Hn u MtM LnN o I a) OO" hnM

rn 1 0 tqnn nO on n noj 000 00

iO H c^.iW i)) o r to > ooi oomi

cnr0Ln 't 1-VoN-L 't I ] cm o t-f "-i 00 rn o

Mc-Ln lrt oI o" c e: o coc:--C-- CD>-o 0CH) O O-tOr-
n-'mLOm 'ocr--oo t' vrC-LnC fOc-ic-i 0r-itflLflm

00000 00000 00000 0000 00000

SLnOr OONOO OOC %OO Ln oCO r00 l OMto

) r-n lO- 00 0 't\00 r- rI mmC)D .cr-I atI -N-C otnC CoDccLn

r- L'- i- L n- 0 0 0 0000 r0 f tt OOr-'C' -tl


I Ln \0
mo3 co oo
mammS i-l ma

N 0 00 -
in ) -N
Ln000 N
000 l C
N00 Cl ) 0 C


A: 000 C
I- 0 C C C
LOA n A t

,-I --I I- -11-

in It n o cNN

o Ln N 0

0 C n I

r- Ar Aoo Ar

In 0n C N %0
C N 000
A1 nr- N N c

C-n In In c N-


N In N-C 00


N In 0 ILn O


0 r

000 A -1

S000 in 00

00 A00 C
0 C 0 C In

In I-n In C -

00-1 --1 0 0

0n In I r- C



CD 0 In 0
SIn CD1 oo

In C(l -1 NO 0)

000'-3 A


000 C In
V~- 00 r-- oI

N N In In 0

C L In In In
0 i-~I *^ltl01
C^ -l0 *0*^
m tl^"LUr^ o
0 0 0 -1 r-

0 N It In 0
00 00 tn In 00
00 I 0)

In InM C C


D0 C -1 C

SN 0 C 0

At tN N0 oo In
o 00 000 C 0

0 % 0 N- C0
-C; I 1C C N

SC C0 00
00 N N C


N> In O In N

1) oo \D1

SC 00 In 0 N

00 C In
N 0lC0
00 InC C



00 0-- -1
0o 0 r 0 A

o CD -i o
LO In LOn Ln


0 0 00 I n
0 0 0 CIn

rL n *t- C ,-l -I 1i-

N N 0 n I 0

0 AD 0 C A
00 000

00 N 00

In i0 \D 0

r- r-- V- r-


L0 In 0 N

SIn In 0
N1 0 N- No 0
- AmA 00

0 r- 0000

Ar- -I- 00
N N 00 In
N~-- N N N

0 mC C
tC n C0 C

O t 'td- L)

CD 0 In C


0- CT In C

In 0H C,. 01

O 00 C

nH 1 N 0


C 0 C 0

Lnm Vd


CD oo oo -I o M OG- cOcoi- Or-li kD f ')O -fn tot
o o omCD Nm ooVcc ft oo LNfo N"0 oo o o E

t C -Nr-Nr-- N- loCo o N-o r- oo 000 0 ) cMcC

00000 0 00000 0000 000 0000


0oLOCUOOLJ) 'D 'D0O-LftLoflnCr t LofLnc"3O-tfl 0o0oLfC %
0D"0c-1Ln L- Oco OLnO DCi0 to) o td- (co c c
ONCO~ 00000 0 f tl00N- C
ur tL) n Ln ZO o-mo Ln n LC'Ln C Ln n oot 00m cC


r-rLo-r-r4 "r mtrC r4 -rq rH,- 0

0000000 t) O OC'Cl"-4 LO r 001)-0"

C0o Cr- C 'o N c'tf OtiOND N N N)lo C)
m mH N w m m m m c)ON m
cNNo C)C CN] CN NoNC)C'-C) C C'cJ CNNC N MC '-Cltfl


oa e <

o" o o oo C) oZa
-N-j4 0] N CO LOO-C to- U

-0000 N N-L tMfl ONOM 000001
o +

o ci -=) iD L -It loo r-- r n V -lo ^ o t- 00 c 0 C o

N- cCOCOC CO t0 C OOM000- OO 0 ONO) 0 N]G
O0O 0 N]OCO )OfN- 0N 0
ooi c4

0Or-COCO C LC-OCdbOOO Ci0OCON 00-70-0
vr- rN-'00D '0 LO,-Ir0<0l rqOCflO 1Ot)O00 U

00000 0 00000 00000 0000 Cl)

) o r L^ ti D ")0C 0 0 C 0 0N 0 00 1 0-lrN

C) OCON0 Ci 00-0t rCOOLOCO oC00OO U)

*O i0 L 0

1t3 i-


S ico -O cN m cI m r) r 't i "- t--

00000 0000 0000 C0 00

00 0 0D 01N t0
10 k cc r-- 71 M -1 0) ft--n co 0) ico Lo co0

S 0n tIn Jn t t v-J tLn- tnC K' W


cooco coo- coornco Uo-o Noco
H +

Com 0 n o L co o- fLOm co n t--i0
0m D tO 1if -0) 't o N)
C OOCSO O N iOC 00 00

0 o --CI 00 Np NN0 q0C0 -o OO
cl Lf

Sr rM- rc- cce GO
C) oo or^ iio cNc* of Doi- cOD- oi ~

w Dv-c"1 I r- M -

Co rD D o D cD " Ln Loo co o i-ico Lo CD

pi, ~ ~ ~ ~ ~ ~ ~ r- pQr O[~'D1- \ ^ *-\> LIC OC -I F't

t- 1- 00 r-N O O4
r- 00 0 00 00

N- C- 00 00 -0

tCl tzl 0 C^] r^ CM Cn

MO tnl Cl

0000 00 0

cq LO "t cNi CT rC- '0
C L- e) -1 cN C

00000000 000000

N- 000N ONN
I oo -4 c- -4


in Lo Uo I 'or -I


-q oC r- oo t():r
,-4N-00 MC0

C 000 i O L-O

0000 000

- r- O-1 -

O i LO o c -i

-0 r-

N- CN N-

00 00

t -4l

00 CO -0

00 1-1 N

N \0 0

00 C:00

N -0 -0

r- No- 1
-I Ln -C)




00 0 -
- Lo Nt- -t

rLO i- co
Nr 0LO

NH- 0)0

CM N- N0 -
N- 00

00 00 00 00

N-N- O- 00

(N L M

On No

-0O -0-0
,- OO

w 0

u 4-

., U-C


N- a) CO)

LO 0- 4J 4O

,t l Le)Cd

S0 -H
00 d
-I0 0 4


H 0*-

0.C 0 0
-)0 04-J
4J V) OL

0-4 4-1 0O
4-J C C4

Q) V 0)L 4-' 0
00 CO0

*4-, t CO L4-
N 0i >

H 4-) j 0
;a4O 0 4 -I1

'C0 4- 0 4-4
O ) -1 0)
tn C cl ,

4J 00 U)O~0'
C-) l C4

CO j 0 rCD
0) -4 4- 0 '
V, 0 z 00 co -4
CL:) 0)4

0 $ -4 0) 't-
4- )0 0 )*-
0. 0 -4. -
0)00)r C
4-)0E C
cO 0CO.0
u' o~r-I0
o .0' 0
*H Co 0
C3n '4-
-rl ,-4.M
C 0) o ) *
CO 00r-
r-4 O c -
mI.0a r,
<0.a- CO

Table 7.3

Lowell Observatory Data

Object Julian Date1 Phase2 R R-I

BY Draconis

GT Pegasus















































Table 7.3 (continued)

Object Julian Date' Phase2 R R-I

GT Pegasus

FF Andromeda





































8.972 +1.166
8.949 1.173







Table 7.3 (continued)

Object Julian Date1 Phase2 R R-I

FF Andromeda 1953.83738 .25545 8.960 +1.102
.84086 8.950 1.112
.94068 .30305 8.958 1.134
.94282 8.990 1.120
.94769 .30628 8.964 1.133
.95127 8.953 1.124

1954.68981 8.999 1.130
.69374 .65008 9.000 1.150
.70093 .65340 8.999 1.135
.70301 9.000 1.130
.84635 .72041 9.004 1.168
.84999 9.020 1.150
.85324 .72359 9.010 1.158
.85654 9.030 1.150
.96933 .77708 9.004 1.163
.97251 9.030 1.159
.97593 .78012 8.999 1.163
.97951 9.015 1.159

1955.72894 9.004 1.168
.73571 .13025 8.997 1.145
.74184 .13308 8.996 1.149
.74791 8.996 1.164
.82218 9.011 1.140
.82940 .17343 9.021 1.155
.83640 .17666 8.985 1.137
.84271 8.995 1.122
.95648 8.983 1.142
.96418 .23554 9.010 1.133
.97674 .24133 8.988 1.162
.98050 8.995 1.140

1956.73432 .59044 8.981 1.153
.73791 8.964 1.130
.82436 .63194 9.008 1.150
.82951 8.992 1.166

1All Julian Dates are to be prefixed with 244.
2Zero point taken as the Julian Date of the first
observation of the observing run, using accepted
periods as given in the text for each star.

Table 7.4

Rosemary Hill Observatory

J.D. @ Oh UT1 Object U-B B-V V "i,'2

1396.5 LP101-15 +1.00 +1.60 12.90 11.50
1400.5 LP101-15 ---- 1.65 12.92 11.49
1420.5 LP101-15 ---- 1.59 12.91 11.48

1545.5 BY Dra +1.03 +1.24 8.44 7.01
1548.5* BY Dra 1.02 1.23 8.45 7.00
1558.5 BY Dra 1.03 1.24 8.46 7.02

1559.5 BY Dra 1.03 1.24 8.44 7.01
1589.5 GT Peg +1.05 +1.51 11.66 9.95
1600.5 FF And +1.09 +1.43 10.38 9.00

1604.5 BY Dra 1.03 1.24 8.44 7.00
GT Peg 1.04 1.52 11.67 9.94
1631.5 FF And 1.08 1.44 10.39 9.01

1678.5 FF And 1.07 1.45 10.38 9.00
1697.5 FF And 1.08 1.44 10.37 8.99
1741.5 LP101-15 1.05 1.62 12.93 11.52

1774.5 LP101-15 ---- 1.60 12.92 11.50
1782.5 LP101-15 ---- 1.59 12.90 11.49
1803.5 LP101-15 ---- 1.57 12.85 11.47

1830.5 LP101-15 ---- 1.60 12.91 11.50
1831.5 LP101-15 ---- 1.61 12.92 11.48
1835.5 LP101-15 1.04 1.63 12.91 11.50

1839.5 LP101-15 1.05 1.60 12.90 11.51
1884.5 BY Dra 1.03 1.23 8.42 7.03
1889.5 BY Dra 1.02 1.24 8.43 7.05

1918.5 GT Peg 1.05 1.51 11.68 9.96

'Julian Dates to be prefixed with 244.
2Zero point term for "R" adjusted to give equivalent
values to those obtained at Kitt Peak and Lowell
Observatories. This is not a well calibrated term.
*This run was unrefrigerated.
Note: The colors and magnitudes above are nightly means.

Table 7.5

Comparison and Check Star

Object U-B B-V V R V-R R-I

comparison +0.261 +0.980 12.610 12.560? +0.050? +0.448
check' -0.06 +0.39 4.84 4.49 +0.35 +0.21

BY Draconis
comparison +1.339 +1.264 8.252 7.008 +1.244 +0.675
check +1.185 +1.309 7.864 6.568 +1.296 +0.705

GT Pegasus
comparison not well determined 10.685 ----- +0.356
check 8.864 ----- +0.581

FF Andromeda
comparison not well determined 8.836 ----- +0.533
check 10.026 ----- +0.411

'These data are accepted standard
Johnson (1966).

values as given by

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 - Version 2.9.7 - mvs