1910.] NATURAL SCIENCES OF PHILADELPHIA. 321 



not distinguishable until the chromatin loses its staining properties. 



Later this body shows a clearer plasmosome part with the chromatin 

 in a band or in scattered granules, and from this chromatin a chromo- 

 some is formed which has the same valence as the other chromosomes 

 and is indistinguishable from them. Because of its behavior this 

 structure is comparable to similar bodies described by Wilson and 

 may be considered as an equal pair of idiochromosomes. 



3. The spireme segments into the reduced number of chromosomes 

 and by condensation the rings become the dyads and tetrads of the 

 first maturation division. By the Foot-Strobell (1907) method of 

 smear preparations, the longitudinal split could be seen in both post- 

 synapsis spiremes and early prophase and its relation traced to tetrad 

 formation and succeeding maturations. The first maturation division 

 is reducing separating univalent chromosomes, while the second is 

 longitudinal and equational. 



4. The same number of chromosomes is found in the equatorial 

 plate of both first and second spermatocytes and by equal divisions 

 the spermatids each receive similar chromosome groups, so that there 

 is no visible dimorphism of the spermatozoa. 



5. Centrosomes were traced from the secondary spermatogonia 

 throughout the whole development of the germ cells. Aster rays 

 and spindle fibres are well developed and the difference between these 

 is especially well shown in mitotic figures in .4. io. 



6. A chromatin granule enclosed by a clearer area is first found in 

 the cytoplasm of the growth stage. This has been described for all 

 forms and in addition an accessory chromatin-like mass has been 

 figured for Acronycta; whether the chromatin granule really functions 

 as the acroblast of King cannot be determined without further research. 

 The accessory mass is present in only part of the cells and is seen to 

 degenerate. 



7. A precocious attempt of the centrosomes to form a flagellum is 

 seen in three of the species of moths studied and in two butterflies. 

 In Promethea this flagellum has been traced from the early growth 

 stage of the first siDermatocyte through prophase, second spermato- 

 cyte, and early spermatid into developed spermatozoa, while it has 

 also been described and figured for the growth stage of A. io, T. poly- 

 pliemus, P. crcsphontes, and D. archippus. This adds additional 

 weight to the view that the axial filament grows out from the centro- 

 some and suggests that its origin is similar to that of astral rays and 

 spindle fibre*. 





