ni "7 



1910.] NATURAL SCIENCES OF PHILADELPHIA. oLi 



^contractile element may be derived directly from the centrosomal 

 substance. The precocious attempt of the centrosome in the first 

 and second spermatocytes to form a flagellum as has been described 

 by Meves, Henneguy, and in the present work for three species of 

 Saturnids, and D. archippus (fig. 191) and P. cresphontes (fig. 192) 

 lends weight to the view that at least in Lepidoptera the out-growth 

 of the axial filament from the centrosome is comparable to the forma- 

 tion of spindle fibres and astral rays. This suggests a possible rela- 

 tionship with other ciliate cells of both plants and animals, and in 

 comparing the vibratile cilia of Lepidoptera with those of plants one 

 is struck with the analogies which exist: the dark-staining granule 

 at the base of the cilium is similar both in appearance and position 

 to the centrosome of the spermatocytes, while the axial filaments in 

 both cases are similar in appearance, vibratile nature, and the part 

 they play in fecundation. Such centrosomes have been figured and 

 described by Webber (1897) for Zamia and by Ikeno (1894) for Cycas 

 and Ginkgo, though the two authors differ in their interpretations; 

 Webber believing that these are only " centrosome-like," while Ikeno 

 does not hesitate to consider them as true centrosomes similar to 

 those of animal spermatozoa. Ishikawa's (1899) observations that 

 the flagellum of Noctiluca grows out from the centrosomal end of the 

 cell, its substance apparently arising from the central spindle, 

 strengthens this view, while Belajeff's (1897) comparison of the 

 spermatogenesis of Characese, Filicineae, and Equisetaceae with such 

 animal forms as the Salamander shows a close relationship between the 

 cilia of spermatozoids and the axial filament of spermatozoa. 



IV. Theoretical Considerations. 



As long ago as 1885 Rabl concluded that each chromosome is a 

 persisting individual and not a structure formed anew in each genera- 

 tion. Van Beneden (1883) advanced the theory of individuality of 

 the chromosomes by pointing out that there is a constant number of 

 chromosomes for each species, always half this number in each matura- 

 tion division and that the number is restored by fertilization. This 

 does not mean that the chromosomes remain unchanged, but that a 

 chromosome of any generation is the descendant of a particular 

 chromosome of a preceding generation. This view was strongly 

 supported by Sutton (1903) who, basing his theory on his own cyto- 

 logical work as well as upon Boveri's (1902) experimental work on 

 the sea urchin egg, concluded that the chromosomes must be the seat 

 of particular qualities, and showed, as Montgomery (1901) had done, 



