348 GEORGE W. BARTELMEZ. 



during the growth of the oocyte the original long axis might be 

 changed more or less, particularly during the final rapid growth 

 period. The oocyte is free to rotate within its follicle at this 

 time and if it cannot hang down into the body cavity so that 

 attached and animal poles coincide exactly, the long axis of the 

 follicle may not coincide with the original long axis of the ovum. 

 At this time the embryonic axis is already established and so the 

 original relation between long and embryonic axes may be 

 changed (cf. Bartelmez, 1912, p. 288). 



The fact that all the most extreme variations were found in 

 eggs with irregular chalazae or poorly defined long axes makes 

 it certain that difficulties of orientation in the oviduct at ovula- 

 tion are important factors in axis angle variability. Further, 

 if the chalazae are not normally attached it is impossible to make 

 accurate measurements. Eleven out of sixteen eggs which had 

 no long axis or very irregular chalazae were extreme variants for 

 the bird that laid them. It is of course obvious that these 

 factors may partly neutralize each other. This will not explain 

 all the variations observed. We must assume a certain amount 

 of spontaneous variability in the eggs of an ovary. 



The first compilations seemed to show that there are always 

 two modes in the curve of individual variation. These appear 

 in Fig. 3, J, 5 and 6, and the same grouping about two modes 

 can also, be seen in the case of the eggs of no. 4. It was natural 

 to expect that this might be correlated with the heterozygous 

 nature of the bird's egg (see Riddle, 1912, 1914). That is to say, 

 the larger, female-producing, eggs might vary about one mean, the 

 smaller, male-producing, eggs about the other. The data from 

 the 58 young hatched in the course of the study gave no evidence 

 that the first egg was more frequently male-producing and Dr. 

 Riddle tells me that this condition is found only in pure races of 

 pigeons. Since also the size and weight of incubated yolks 

 varies greatly from the absorption of albumen this possibility 

 could not be tested out. 



It is however a matter of great practical importance to see 

 what the chances are that the first and second eggs will have 

 about the same angles. In collecting early stages we usually 

 determine the stage of the second egg from the time of laying of 



