-I T.KMATOGENESIS OF THE CADDIS-FLY. 5/ 



with very good results. The sections were stained in Heiden- 

 hain's iron alum ha-matoxylin or in Flemming's triple stain. 



THE TESTES. 



The testes, a- Yorln -i- (17) has already noted, occur in the 

 fourth and fifth -e-ment of the abdomen. They are small five 

 loltrd -trurtnre- n-tially -urrounded by a mass of fat. 



Tin -. ti\r lob.-- <ir fi.llicles are about equal size but of various 

 shape-. In general the\ are somewhat wedge-shaped, tin- thinner 

 part nl the \\< (!- l.i in-, of course, toward the point of attach- 

 in" -in of the oriMn I U. i.}). In longitudinal section about fifty 

 ippear in cadi f"llile (Fig. 14), the larger one- at it- 

 -mailer end. 



Karh folliele ha- it- ou n wall and the flattened nuclei of th. 

 \\all i ell- |ie.|ueiitly appear in the sections. In addition each 

 i- pio\ided \\iih a \\all. The youngest cells are fart he- 1 

 Iti'in the point ol aiia< hmeni of the testes. In one -ertion. as 

 -hounin li^. i( ntaining spiremes for the first division, 



tei rads, in -i and se< . md di\ i-ions, transformations of the speiina- 

 tiiU. and Inllx de\ duped -perms occur in the order named at ; 

 the -e( limi. The \ min^ i \ -IN show in section, 6-8 spermatot \ tes; 

 the older mie-. \\ here the eell- are smaller, due to the two di\ i-inn- 

 and to the growth of th( ~, show 12-18. There is no del in it r 



/oiiation i.| -piienie-, lir-i divisions, second di\i-ioiis, etc., surh 

 a- \Vili <i\ i'i has loured in Caloptenius femur-rnbrntn, but the 

 m-m-ral lendeiK \ i- for the cysts containing the younger cell- 

 to be further a\\a\ limn the smaller end of the follicles. 'I In- 

 tact that there i- in. definite Donation of su< ve--i\ely older spcrm- 

 atocytes make- it at fir-t confusing in picking out the se(|in m e 

 I de\eloiMueiu. The difference in size and number ol them in 

 the different cysts ho\\ ever makes this possible after a little 

 observation. The -permatid- in developing in the cysts, as is 

 n-ual, ha\e their head- all pointing in one direction as they lie 

 -ide \>\ -ide. 



Tin SIM KMAGONIAL CELLS. 



As tin earl\ hi-t>r\ of the -permagonial cells does not seem to 

 have been worked out on many species of insects a rather full 

 description of that process \\ill be given here. In this insect the 



