53 Edmund B. FFilson. 



The four classes thus formed may be tabulated as follows: 



Primary Class A f (i) 12 ordinary chromosomes, I heterotropic, I large chromosome. 



(14 chromosomes) ( (2) 12 ordinary chromosomes, I beterotropic, I small idiochromosome. 

 Primary Class B j (3) 12 ordinary chromosomes, I large idiochromosome. 



(13 chromosomes) (^ (4) 12 ordinary chromosomes, I small idiochromosome. 



Restating the facts from the point of view of mere size, it appears 

 that class (3) contains no especially small chromosome, class (2) 

 two small chromosomes, and classes (i) and (4) each one small 

 chromosome, the latter being in one case the heterotropic, in the 

 other the small, idiochromosome. 1 



I have not yet studied in sufficient detail the history of this 

 form in the growth-period, which will require additional material; 

 but the main facts are such as might be expected. In the middle 

 growth-period the nuclei show, with great constancy, two unequal 

 chromosome-nucleoli, both of which frequently appear hollow. 

 The larger of these is almost certainly the idiochromosome- 

 bivalent; for in the prophases of the first division it may be seen 

 separating into its two unequal constituents, precisely as I 

 described in Brochymena (Fig. 4, /?, q). At this period the hetero- 

 tropic chromosome is unmistakably recognizable by its size and 

 shape, showing no constriction like that of the other chromo- 

 somes. I believe this to be identical with the smaller chromo- 

 some-nucleolus of the earlier period, but cannot offer decisive 

 proof. 



CRITICAL AND COMPARATIVE. 



The three well-defined classes of chromosomes that have been 

 described in this and my preceding paper differ from the others, 

 each in its own way, especially in respect to their behavior in the 

 process of synapsis and during the growth-period. The most 

 characteristic common feature of the first two classes is their long 

 delayed synapsis, which in both cases is deferred to the period 



*It is probable that additional light will be thrown on this form by further study of the related one, 

 Thyanta custator, which I now have under investigation. The general aspect of the chromosome group 

 in this species is closely similar to that of Banasa, and the first mitosis also shows fifteen chromosomes, 

 of which however three, instead of two, are smaller than the others. The second mitosis differs from 

 that of Banasa in showing always but thirteen chromosomes, and I have not thus far found a heterotropic 

 chromosome in either division. Though I cannot yet speak positively, these conditions seem only 

 explicable under the assumption that two pairs of idiochromosomes are present. From such a con- 

 dition one nearly similar to that observed in Banasa might readily be derived by the disappearance of 

 one of the small idiochromosomes. 



