720 C. E. McCLUNG 



center of the equatorial plate (fig. 57). Only at the point of fiber 

 attachment do the chromatids remain in contact. Separation 

 of the chromosome halves already exists at this time, and move- 

 ment to the spindle poles as simple, straight rods easily follows. 

 The accessory chromosome, being also a dyad, divides like the 

 other chromosomes. Upon these points there is practical agree- 

 ment by all who have studied the Acrididae, so a further discussion 

 is not called for; but whether the two chromatids are sister ele- 

 ments or merely homologues is a mooted question. It entirely 

 depends upon what has happened in the first spermatocyte and 

 the evidence of the earlier generation is much clearer. Accord- 

 ing to my belief, in most of the chromosomes they are homologous 

 elements and are separated, through a segregation^ division by this 

 mitosis. 



2. Stenohothrus type. In this, as in the Hippiscus type, the sizes 

 of chromosomes in the spermatogonium reappear in the second 

 spermatocyte but, in addition to this, also the forms recognizable in 

 the earlier generation are found. Indeed, aside from the haploid 

 condition of the chromosome group, all the relations of the sper- 

 matogonium are preserved. As may be seen from figures 68 to 76, 

 there are, in the Stenobothrus-like form I have studied, always 

 six Hippiscus-like rods in each second spermatoc^'^te and four 

 J- or V-shaped dyads. ^Yhen the accessory chromosome is pres- 

 ent there is added a fifth chromosome with non-terminal fiber 

 attachment. In Stenohothrus curtipennis Davis reports a simi- 

 lar constancy of forms and his conclusions are borne out by the 

 work of ]\Ieek and supported by the less detailed studies of de 

 Sinety and Gerard. It seems established be3'ond question that 

 organization is so exact that even the form of the chromosomes is 



' In this paper and others I have used the term 'segregation division' synony- 

 mously with 'reduction division. ' It seems to me a preferable term with our present 

 knowledge of cellular processes, since it indicates the nature of the separation in 

 concordance with our understanding of the composition of the tetrads out of 

 elements derived from the two parents. There is no 'reduction' in Weismann's 

 sense, only a segregation of paternal and maternal derivatives. Of course this 

 involves the conception of such a biparental composition of the tetrads, but prac- 

 tically the same idea is at the bottom of any 'reduction' scheme and so it would 

 not hold more against one than the other. 



