The Spermatogenesis in Pentatoma up to the Formation of the Spermatid. 77 



elusions are justified in applying our definition of chromosome to the 

 chromatin elements of the reduction divisions, in those cases which 

 have been most thoroughly described (Ascaris, Copepoda, Caloptenus, 

 Gryllotalpa, Lumhricus, Ophryotroclia, Salamandra, Selachii, Pyrrho- 

 coris and Pentatoma). In most cases the number of chromosomes in 

 the 1st division is half the normal number (Ascaris, Copepoda, Gryllo- 

 tdlpa, Salamandra, Selachii, Pentatoma). In the remaining cases the 

 number of chromosomes in the 1st division is the same as the normal 

 number (Caloptenus, Lumhricus, OphryotrocJia and, according to Hen- 

 king, PyrrJwcoris). There would appear to be no cases where double 

 the normal number of chromosomes appears in the 1st division: the 

 only case which would seem at first sight to come under this category 

 being Caloptenus, which must be relegated to the preceding one. That 

 is to say, if we apply our definition of a chromosome to the reduction 

 divisions, there has as yet been made known no undoubted case of a 

 doubling of the chromosomes, and those who have assumed such a 

 doubling (as Weismann and some of his followers), have reached this 

 conclusion by applying an unusual definition to the idea chromosome. 



Two modes of chromosome division are distinguished, the equation 

 and the reduction, as first applied by Weismann. The equation divi- 

 sion is a longitudinal division of a chromosome, and is supposed to 

 separate from one another ids of equal value. Two kinds of reduction 

 divisions are distinguished, both of which are supposed to result in 

 the separation of unlike ids from one another : it may be a transverse 

 division of the chromosome (which may be termed reduction in the 

 sense of Rückert); or the reduction division may consist in the 

 separation of whole chromosomes from one another (reduction in the 

 sense of Weismann, for which the term "elimination" might be sub- 

 stituted). 



I am inclined to think that the mode of division of the chromo- 

 somes may be in some way, by a hitherto unexplained law, a function 

 of their form. It would appear that thin, ribbon-shaped chromosomes, 

 like those of Salamandra in the reduction divisions are more apt to 

 be longitudinally split; and that thicker chromosomes, of a massive or 

 spherical shape, more frequently divide transversely. In any case, it 

 is only in a very thin chromosome, formed of a single layer of micro- 

 somes (Weismann's ids), that by a longitudinal division each micro- 

 some becomes halved. While by transverse division of any chromo- 

 some, or by longitudinal division of those which are formed of a thick 

 mass of microsomes, one or only a few of the microsomes (and then 



