GENETICS: W. E. CASTLE 
253 
out size or devoid of genetic variation in size? No race of animals has 
yet been shown to exist which is devoid of genetic variation in size, — 
Jennings' supposed demonstration of such a condition in asexually 
reproducing Paramecium having been disproved by the work of Calkins 
and Gregory and his own subsequent work on Difflugia. If when all 
discoverable Mendehan factors are wanting size still varies genetically, 
a Mendelian explanation of size inheritance proves itself inadequate. 
There also exists a possibiity that hantamness in fowls involves discon- 
tinuous variation such as does not occur in ordinary size differences. 
For Phillip's results with ducks do not show such segregation of the 
extremes of size as Punnett and Bailey record, though his crosses were 
made between races of ducks both very pure and very different from 
each other in average size. What Phillips observed was a blend with a 
slight increase of variability in F2 but without evidence of complete 
segregation or recombination of size factors. 
Some excellent material for the study of size inheritance has become 
available to the writer in certain very pure races of guinea-pigs differing 
widely as to size and it is the purpose of this note to describe briefly the 
more important results obtained and their possible significance. Atten- 
tion is invited to the nature of the growth curves observed for the races 
crossed and to non-genetic as well as to genetic factors affecting size. 
Three distinct and unrelated races were used in the crosses to be 
described. First, a wild race obtained at Arequipa Peru in December 
1911, and identified as Cavia Cutleri Bennet. This has been bred for 
three generations in captivity at the Bussey Institution and has shown 
itself very uniform in size and other characters. Second, a race of 
guinea-pigs which we may call race B, bred distinct for many years at 
the Bussey Institution, very uniform in size, and varying as regards 
Mendehan factors only in respect to the color factor. Some animals 
of this race are black, others are albino. Both sorts are alike as regards 
other characters. A race C, also long bred distinct at the Bussey 
Institution, is of about the same average size as race B. It as well as 
race B has been used in crosses with C. Cutleri, but the hybrids from 
these two crosses have not been interbred, though the data concerning 
them have been combined for statistical treatment. 
Figure 1 shows separately growth curves for the two sexes of race B 
and Cavia Cutleri. These curves have been obtained by combining the 
individual growth curves of several different animals reared in captivity 
and weighed at intervals of one or two weeks. They represent averages 
and the curves have been smoothed somewhat. It will be observed 
that the young of Cutleri grow rapidly for the first 40 or 50 days of their 
