SPERM ATOGENESIS OF THE CADDIS-FLY. 6 1 



follicles, which indicates, of course, that five divisions have oc- 

 curred. 



After the la-t division, the cells of these mature cysts grow 

 ;md the nucleu- becomes much larger in proportion to the size 

 of the cell. Fig. 8 shows the size of the cyst immediately after 

 it ha- acquired it- full number of cells and Fig. 10 shows a cv-t 

 after tin- gnmth period in which the cells are in the -pireme for 

 the reduction di\ i-ion-. There are therefore two growth periods: 

 after the la-t primary spermagonial divi-ion and the >econd 

 i In- reduci ion di\ isions. 



Tin REDUCTION DIVISIONS. 



Alter the l.i-t -pel in.e^oiiial divisions the growth i- apparently 

 very rapid. The nucleus is very large, occupying almo-t the 

 entire cell and the chromatin is in the form of regular little pi, 



Fij it. The remain- of the spindle of the last >permagonial 

 division lie- near tin- IUK leu> in the form of an <>\oid body, the 



ii-iikern. 



Apparently the fn-t -tage in division shows the little pieces 

 ol i hioinatin draun out until they resemble ehroino-, ,nn-> while 

 the mi, leo lu- become- oval or spindle-shaped big. 12). The 

 pieces o| rhiomatin lu-comc united by a slender connection and 

 -eem to -pin out and lo-- their identity. The niu leolu> in the 

 meantime becomes more and more elongated Fij 15). 



The chiomatin n\\ -e.-ms to go into a \u^ -lendi-r >|)in-ni, , 

 the ma-- ol the thread- King at one side <t" the nucleus and oc- 

 eii|)\ini; about one half of it. This is apparently the stai^e of 

 >\ nap-i-. The -i rand- are very fine and delicate and as the nuclei 

 are not -, . \,-r\ lar^e. it is impossible to make out an\ pairing 

 ol the threads, if such occurs. The nudeolus lies as a long drawn 

 out l>od\ either in thi> tangle of the threads or on its surface. 

 ' ', a-ional loop- >tick up out of the denser clum]> but they are 

 too -mall to follou or to make much out of. After tlu-e -tages, 

 \\hich are a|)parentl\ -hort a- they are not nunierou-, the chro- 

 matin again till- the entire cell in the form of the spireme. This 

 spireme gradual!) become- -horter and thicker. This stage is a 

 \er\ Ion- one as i- -hown by the fact that in sections where any 

 di\i-ion- ,u all occur, about half of the cells are in this stage. 



