ZOOLOGY: A. J. GOLDFARB 
241 
tion during the earlier months of the experiment may be largely the re- 
sultant of other interrelationships. Consider this possibility in detail. 
Let Tye^, Tye^, Tye^, . . . Tye^^ be the correlations between percent yel- 
low in the twelfth month of the contest and egg production in the first, 
second, third . . . twelfth months respectively. Further let 
re e . . . Te e be the correlations between October and November, Octo- 
"12 J 12 11 ' 
ber and December, October and January . . . October and September 
egg productions. These constants have been shown to be positive 
throughout, indicating that birds excelling in egg production in October 
gave on an average higher productions in every other month of the year. 
The application of the well known partial correlation formula for 
one variable, 612, constant results in very material reductions in the 
values of fye^, rye^, rye^, . . . rye^^. Thus the values of Tye^, rye^, rye^, . . . 
Tye^^ must be in large part the resultants of Te^^e^ and fye^^, fe^^e^ and rye^, 
Te^^e^ and Tye^^ . . . Te^^e^^ and Vye^^. 
A discussion of the biological and biochemical literature, and a de- 
tailed statistical treatment of the data are appearing in Genetics, 1917. 
* Ear lobe color has been measured in units of 5% range by means of the color top. Only 
yellow and white sectors were used. 
2 All birds entered the international Egg Laying Competition at Storrs, Connecticut, 
in November of their pullet year, and remained until the end of October of the following 
year. Pigment determinations were made near the end of October. 
3 When the constants for the comparable months in the two years are considered in 
comparison with their probable errors, there is not a single difference which can be considered 
significant. 
^ The biological inference to be drawn from this result would seem to be that the egg 
production of a recent period influences very profoundly the concentration of yellow pig- 
ment, so that there is a very rapid decrease in yellow pigment for each additional egg laid 
up to a certain point, beyond which the body pigment is relatively little reduced by extra 
egg production. Thus for October, the change in pigmentation is to be described by a curve, 
not by the slope of a straight line. The change in pigmentation is not proportional to egg 
production, but at first is very rapid and then falls off. 
^ In collecting these data a bird which laid on the day the pigment determination was 
made or on a later day within the month was considered to be laying, and was recorded in the 
zero class, i. e., no days since laying. If she laid on the day before the record was taken 
but not later she is recorded as one day since laying, and so on. 
VARIABILITY OF GERM CELLS OF SEA URCHINS 
By A. J. Goldfarb 
COLLEGE OF THE CITY OF NEW YORK. AND DEPARTMENT OF MARINE BIOLOGY. CARNEGIE 
INSTITUTE OF WASHINGTON 
Communicated by A. G. Mayer, February 6, 1917 
As a basis for an understanding of the changes in aging germ cells, 
it was necessary first to ascertain the normal variability, i.e., the varia- 
