278 THE UNIVERSITY SCIENCE BULLETIN. 



Diabroticas to have from one to five supernumeraries and 20 

 per cent of the males in Cetdhophihis to have one or more 

 supernumeraries. Her conclusion was that these bodies were 

 probably related to X-chromosomes, descended from frag- 

 ments of X-chromosomes that had arisen through irregulari- 

 ties in the maturation division at some previous time. Wilson 

 regarded his supernumeraries in Metapodius as duplicates of 

 the Y-chromosome, i. e., of the smaller member of the idiochro- 

 mosome pair. No such Y-chromosome has been described for 

 Acrididse or Locustidse. In addition, the irregular behavior 

 of this supernumerary body (s) with reference to the access- 

 ory chromosome (5x) , in going either with it or not to one 

 pole in the first spermatocyte, is evidence against its being a 

 Y-chromosome. This was also true of the Diabroticas and 

 Ceuthophilus. It differs, however, from the latter two cases 

 in this : that no cell has been found showing this body dividing 

 in first instead of the second spermatocyte division, as was 

 true in a few cases in the Diabroticas and Ceuthophilus. Other- 

 wise it behaves very much like the supernumeraries of these 

 forms. It seems warrantable to consider it as belonging to the 

 X-chromosome rather than the Y-chromosome category. 



If this body is in reality an accessory chromosome, and one 

 that has lost from the distal end one-fifth or one-fourth of its 

 length, and if it is also to be considered functional in so far as 

 the portion present is concerned, we may regard it as throwing 

 some light upon the question as to how much and what portion 

 of the accessory in the family Tettigidx is concerned with sex- 

 determination. In the first place, that the sex-determining 

 genes are probably limited to only a portion of the accessory 

 chromosome in the genus Tettigidea, at least is to be inferred 

 from a comparison of the relative lengths of the chromosome 

 in the different genera of the family to which the genus belongs, 

 the chromosome varying from eight in Acridium to fourteen or 

 fifteen in Paratettix and eighteen or nineteen in Tettigidea. 

 This supposition is of course based upon the assumption, first 

 of all, that the X-chromosome is to a large degree similar in 

 constitution throughout the family. That this is true w^e have 

 reason to believe from the fact that these genera which appear 

 to be so closely related taxonomically possess chromosomes 

 (both autosomes and allosomes) which have a very great de- 

 gree of similarity in number, form, and relative size (Robert- 



