308 PROCEEDINGS OF THE ACADEMY OF [April, 



From this stage the centrosomes which have been followed through 

 every step of development show unfailing persistence and regularity 

 in movement, in form, and in division. This observation, like that 

 of Conklin (1902) for Crepidula and Paulmier for Anasa, supports the 

 view that the centrosome is entitled to the rank of a permanent cell 

 organ. 



In the succeeding growth stage the centrosomes surrounded by 

 faint astral rays are seen to separate ; then they divide : one pair moves 

 towards the nulcear membrane, the other pair lies upon the cell wall 

 (fig. 25). In preparations which have been somewhat deeply stained 

 cilia of considerable length are seen to have grown out from each 

 granule and extend into the lumen of the cyst (figs. 24, 25 and 29); 

 these correspond to the ciliated centrosomes figured by Sieves for 

 Pygcera (1897) and by Henneguy for Bombyx (1898), differing only 

 from those shown by Meves in the shapes of the centrosomes. which 

 in Pygcera are described as hooked or Y-shaped bodies. In C. pro- 

 methea these are plainly small dumbbell-shaped structures and are 

 exactly like the centrosomes described by most workers, differing 

 only in the possession of the flagellum. 



While the chromosomes are passing through the prophase of the 

 first maturation division, the paired centrosomes, enclosed by a small 

 centrosphere, move to opposite sides of the nucleus where they appear 

 as dumbbell-shaped bodies surrounded by a clear archoplasmic zone 

 and short radiating fibres (figs. 31 and 32). As the nuclear membrane 

 disintegrates, the astral figure increases in size, its fibres extending 

 outward into the cytoplasm and inward towards the middle of the 

 cell, and as division proceeds the astral raj^s increase in length until 

 they seem to press against the wall of the cell. I have never observed 

 a flagellum going from the centrosome at this stage, though such has 

 been figured and described by Henneguy. After completion of the 

 metaphase the centrosomes divide and appear in the telophase of the 

 first maturation division as two separate bodies, each surrounded by 

 an archoplasmic mass ; the centrosomes migrate to opposite sides 

 of the nucleus in an axis at right angles to the first maturation spindle. 

 Just at this time a flagellum is seen to grow out of one of the centro- 

 somes (fig. 48), this flagellum, though somewhat longer, is similar 

 to those described for the first spermatocytes. 



One other structure in the cytoplasm which is of interest is the 

 "chromatin granule" first seen during the growth period (fig. 25). 

 These granules, either single or dividing, stain like chromatin, are sur- 

 rounded by a clear zone, and are traceable through succeeding stages'; 



