THE SPERM ATOGENESIS OF AGALENA N.EVIA. 133 



from the outer wall leaving a clear area between it and the chro- 

 matin mass. Wagner described the final form of the spermato- 

 zoon as being rod-like and was criticized in this by Bosenberg 

 who suggested that he was misled by seeing cross- sections of the 

 disk-like sperm. This cannot possibly be the explanation for 

 not only are all of the spermatozoa in the distal portion of the 

 ducts and in the pedipalps in this elongated form but they are 

 larger than when in the proximal portion of the ducts or in the 

 lumen. This fact precludes the possibility of their being cross- 

 sections (Fig. 59). Although they remain coiled they seem to 

 lose the compact structure, which they assume before leaving the 

 cysts, and regain their original size. This reexpansion of the 

 heads seem so improbable that I was only convinced of its truth 

 after making numerous camera lucida drawings with utmost care. 

 The difference in size and form can be seen at a glance by com- 

 paring Figs. 56 and 59 which were drawn with the same magni- 

 fication. In all probability the spermatozoa first uncoil after they 

 are passed through the slender ducts of the receptaculum sem- 

 inis into the oviduct of the female spider. 



When one considers that the tail is attached to the lower, an- 

 terior margin of the relatively heavy, crescent-shaped head, it is 

 difficult to see how its lashing movements would propel the head 

 in a straightforward direction. Bosenberg has made considerable 

 study of the movements of spider spermatozoa and he holds that 

 the chief propelling power is found in the twisting and bending 

 motions of the head itself while the tail, by its lashing move- 

 ments, may act as a steering-organ. 



Sequence of Divisions. 



In the maturation of the male germ cells it is generally con- 

 ceded that two kinds of division occur a reductional and an 

 equattonal one but in regard to their sequence there has been 

 much difference of opinion. At a time when the majority favored 

 the view that the equational division occurs first, Montgomery 

 {'04, "05) maintained that the first division is reductional and he 

 strongly emphasized the importance of determining the origin of 

 the chromosomes in synapsis as their mere form in post-synapsis 

 -stages is often misleading. To-day perhaps the majority of 



