CRITICISMS AND CONCLUSIONS. 55 



"with absolute certainty" 12 chromosomes, and in many other instances 

 approximately 12. Now, however, apparently without any additional 

 material, he (1907, p. 512; 1908, p. 248) counts 16! Gerlach (1906, p. 23) 

 expresses himself as emphatically agreeing with Sobotta in his early 

 statement that the number is 12, he (Gerlach) having repeatedly counted 

 12 in both the first and the second spindle. Lams et Doorme count 

 the same number, 12, in two polar cells; but we have shown (p. 42) that 

 the number in the polar cell has no significance. Kirkham (1907b, pp. 

 74-78) likewise affirms that there are 12 chromosomes, and in those cases 

 where there are obviously more than 12 bodies he explains the higher 

 number as being due to the precocious division of some of the chromo- 

 somes. Nevertheless, in Kirkham's own preparations, which were so 

 generously loaned to us, out of four normal ovarian eggs in the stage of 

 the first spindle there were three cases in which we could count 20 with 

 certainty, and in the remaining one 17. 



ACHROMATIN. 



Gerlach (1906) and Sobotta (1908, p. 508) are the only writers on 

 the maturation of the egg in mice who give any opinion as to the precise 

 origin of the fibers of the first spindle. These they think arise from the 

 linin network of the germinative vesicle. But this seems improbable in 

 view of the fact that there is a stage before their appearance in which 

 only shreds of the linin network are left, while most of the vesicle is filled 

 with a clear fluid. It is possible that the linin plays some part in the 

 origin of the spindle; but, as has already been suggested, other parts of 

 the nucleus, including the nucleolus, are the more probable sources. 



Tafani has pointed out that in its early stages the first spindle in 

 ovarian eggs is short and fat, a condition we also have found. Sobotta 

 (1895, 1899, I 9°7) figures in a diagrammatic way the spindle with sharp 

 poles, the fibers converging to a point. Lams et Doorme (1907, p. 274) 

 say the fibers converge more or less to a point. Kirkham figures the 

 shape of the first spindle as elliptical. 



According to Sobotta (1907) the largest spindle is 30 to 32 micra 

 long and 20 micra broad. The largest spindles we have found have the 

 following dimensions: 29.5 micra in length by 11 in breadth, and 22.6 

 in length by 14 in breadth. From Sobotta's paper of 1899 it must be 

 inferred that the size varies. The statement of Lams et Doorme (1907, 

 p. 275) and our own observations accord with this inference. Gerlach's 

 statement (p. 10) that the size depends in the main on the size of the 

 germinative vesicle can not be accepted as demonstrated, for the spindle 

 is not a result of the metamorphosis of a network confined in a rigid 

 vesicle; besides, the membrane of the vesicle has nearly disappeared 

 when the spindle is first differentiated. 



Sobotta described the spindle fibers in 1895 (p. 51) as fine, wavy, 

 and branched; in 1907 (p. 508) as wavy with slight thickenings. His 

 latter description applies to the early stages of the first spindle, for later 



