42 THOS. H, MONTGOMERY jr., 



1st spermatocytic division marked irregularily in the form and size of 

 the chromosomes, between the elements of the same nucleus as between 

 those of different nuclei. This results in the formation of 7 chromo- 

 somes which, as a rule, differ from one another very noticeably in 

 volume , and to some extent also in form , though the differences in 

 form are more characteristic of the earlier prophases, before the 

 definitive dumbbell shape has been attained. Their form at the loose 

 spirem stage varies greatly, and no two can be found showing exactly 

 the same outline (Figs. 124—136). They are now as a rule rod-shaped 

 or irregularly U-shaped or V-shaped ; and in those destined to segment 

 into definitive chromosomes, and such are usually the longest loops, 

 the transverse constriction usually appears at the bend of the U or 

 V. Even those which already have the value of definitive chromo- 

 somes, i. e. those which do not divide further until the metakinesis 

 of the 1st division, usually show a transverse constriction at the point 

 of bending. This fact is important, since it shows that as early as 

 the commencement of the loose spirem many of the chromosomes 

 already show the transverse constriction, which they all have just 

 before the disappearance of the nuclear membrane. A glance at 

 Figs. 124 — 146 demonstrates conclusively how early this transverse 

 constriction makes its appearance, though different chromosomes vary 

 from one another in regard to the time at which it first appears. 



The loose spirem cannot be sharply demarcated from the dense 

 spirem nor from the end of the prophase; it may arbitrarily be said 

 to terminate when the 7 chromosomes have attained their definitive 

 dumbbell shape. 



In many nuclei, besides the rod-shaped and bent chromosomes, 

 are found ring-shaped ones (Figs. 122, 125, 126, 129, 130, 132, 135, 

 143 — 147), the origin and valence of which remains to be discussed. 

 The presence of such ring forms is not characteristic of a particular 

 stage; in the majority of nuclei where they occur, only one is found 

 to a nucleus. I estimate that they do not occur in more that one- 

 fifth or one-fourth of the cells at this stage, so that their presence 

 must be considered rather exceptional. It is, of course, necessary 

 to determine whether they have the value of univalent, bivalent or 

 quadrivalent chromosomes, and whether their form might be due to 

 processes such as produce the chromatin rings in the heterotypic 

 mitoses of the salamander spermatocytes, or those of Gryllotalpa. 

 Such questions are to be solved in part by the examination of nuclei 

 in which all the chromosomes lie in the plane of the section, and a 



