524 The Ohio Naturalist. [Vol. VI, No. 7^ 



ing of granules in fours may be, it is not necessarily a precourser 

 of the reduction division. Such groups of four granules as Brauer 

 shows (fig. 22-2-i) are frequently very abundant in the nuclei of 

 the wall cells of the uterus which are not, of course, in prepara- 

 tion for a reduction division. He has several figures (35, 37, 

 41, 42) in which the doubled loops shown in my fig. 6 are very 

 plain. He does not, however, follow the gradual approxima- 

 tion of the sides of the loops but supposes them to straighten 

 out into a single semicircular band which by transverse division 

 forms the two tetrads. During all these stages he supposes that 

 the spirem is composed of the doubly split granules of the early 

 prophases, believing, doubtless, that his inability to see them 

 was due to the very unfavorable positions which such objects 

 would inevitably assume. He does, however, show in small por- 

 tions of figures 34, 36 and 41, places where the spirem is repre- 

 sented as composed of three or four strands instead of two which 

 the present writer has invariably found. Beyond these points 

 there is no greater difference in our observations than is pro- 

 bably due to differences in the sex cells of male and female 

 animals. 



Montgomery (6) has pointed out that the tetrads are of un- 

 equal size. My own studies have not been carried carefully into 

 the maturation mitoses where Montgomery made his observa- 

 tions but what I have seen of these stages tends to confirm his 

 statements. The earlier stages also offer strong confirmatory 

 evidence of their truth. As has been mentioned one of the tet- 

 rads is almost always slower in its formation than the other, be- 

 ing derived apparently from longer more contorted segments of 

 the spirem thread. His contention is that there is always a con- 

 jugation of similar chromosomes to form a tetrad. This would 

 seem to be correct in the main but it appears to be not without 

 exceptions, see fig. 8. 



SUMMARY. 



The foregoing observations seem to show conclusively that 

 the tetrads in the eggs of Ascaris megalocephala bivalens arise 

 not by a double longitudinal split of the original spirem thread 

 but by a folding of adjacent segments together (conjugation of 

 univalent chromosomes) together with what is believed to be a 

 single longitudinal split. The two split loops which form the 

 two tetrads appear very early in the contuinous spirem and in 

 their later development simply break apart, shorten, thicken, 

 and straighten out till the tetrads are formed. 



Since of each tetrad only one component chromosome re- 

 mains in the ripe ovum, there is a Reducing Division in Wesi- 

 mann's sense In' which paired chromosomes are separated from 

 each other in the egg and the hereditary characters transmitted by 

 the chromosomes, thereby modified. 



