Studies on Chromosomes. 517 



I believe that in this form, too, the m-chromosomes are fre- 

 quently recognizable as condensed separate bodies in the growth- 

 period; but owing to their minute size it is difficult to be sure of 

 this. In any case, in the period just before the disappearance of 

 the nuclear membrane they are quite distinct from the "acces- 

 sory," which is, as in Alydus, immediately recognizable by its 

 size (Fig. 3, g). From this period, as in Alydus, the latter body 

 may be traced continuously backward into the growth-period. 



The foregoing facts, observed in Alydus and Archimerus are 

 in close agreement with Montgomery's results on Protenor, 

 differing only in that the condensation of the m-chromosomes 

 takes place somewhat later. 1 In Anasa the condensation of these 

 chromosomes from the diffused condition takes place still later; 

 and this, combined with the fact that the "accessory" cannot be 

 certainly distinguished from the other larger chromosomes by its 

 size, renders the question more difficult of solution than in Alydus, 

 though I believe the result is equally decisive. In Anasa, as in 

 Alydus or Archimerus, the small central bivalent of the first 

 equatorial plate is formed by a very late conjugation of two 

 separate microchromosomes that only come together as the spindle 

 forms, precisely as Gross describes in Syromastes. Of this fact 

 no doubt is left by the study of a large number of preparations 

 that show every stage of the process, step by step. In the late 

 prophases, just before the nuclear membrane disappears, the nuclei 

 invariably show twelve separate, condensed, intensely-staining 

 chromatin-elements (one more than the number of chromosomes 

 in the first mitosis) in addition to one or more pale rounded 

 plasmosomes with which the chromosomes cannot for a moment 

 be confused. Ten of these are larger bivalents which have the 

 form of quadripartite tetrads or dumbbell-shaped bodies. The 

 remaining two are much smaller bodies, irregularly ovoidal or 

 frequently more or less distinctly bipartite (m, Fig. 2, ^, /); they 

 may occupy any relative position. As the spindle forms, the 

 microchromosomes lose their bipartite form, assume an evenly 

 rounded ovoidal shape, and conjugate at the center of the equa- 

 torial plate to form a small dyad-shaped bivalent (Fig. 2, -*) 

 Without fusion the two halves are then immediately drawn apart 



'In Alydus pilosulus this author believed the /n-chromosomes, as usual, to be derived from the 

 large chromosome-nucleolus. 



