The Germ Cells of Aphis. 321 



it is impossible to distinguish the individual chromosomes in the 

 spindle; they are crowded together and often are so united in 

 metaphase as to resemble a spireme in one plane. 



IV. SPERMATOGENESIS. 



Nearly all of the material for the study of the spermatogenesis 

 w^as obtained from the CEnothera. On the rose the young males 

 are never met with in large, numbers w^hile on the CEnothera an 

 abundance of all sizes may often be found on the still-blossoming 

 tips of the flower spikes, while the sexual females are scattered over 

 the leaves and stalks. Here also were found a few mothers of the 

 males. The few young males from the rose showed the same 

 number and relative size of chromosomes. 



The testes, as seen in dissections, have 6 lobes corresponding to 

 the 6 ovaries in each group. The lobes are much larger than the 

 ovaries, and as the spermatogonia are somewhat smaller than the 

 oogonia, the number of mitoses leading to the spermatocyte must 

 be several more than in the case of the oogonia. Only spermat- 

 ogonial divisions are found in the embryos, and the last such 

 division often, if not alw2f}s, occurs after birth. Fig. 28 shows a 

 resting spermatogonium ; Figs. 29 and 30 spermatogonia just before 

 division. In Fig. 30, 9 chromosomes of characteristic form and 

 size can be recognized, one of the smallest not being visible. The 

 equatorial plate, as also in most somatic mitoses, is too crowded 

 for distinguishing either number or form of chromosomes. 



The resting spermatocyte of the first order (Fig. 32) does not 

 differ materially in appearance from the resting spermatogonium 

 (Fig. 28) and indeed can often be recognized only by its relation 

 to later stages. There is no evidence of so long a growth stage as 

 in most forms. Closely associated with the first maturation 

 mitosis and probably immediately preceding it are found the 

 stages shown in Figs. 33 and 34. In Fig. 33, a, 8 of the 10 chro- 

 mosomes are to be seen scattered through the nucleus; in b, c and 

 d, chromosomes of the same form and size are seen paired longi- 

 tudinally. Fig. 34 suggests the synapsis stage described for manv 

 insects. This stage is of much more frequent occurrence than 

 that shown in Fig. 33, and appears to follow that, and immediately 

 precede the first spermatocyte division which is shown in Figs. 

 35-39. Fig. 35, <7, h and c show the 5 chromosomes in the equa- 



