162 CHROMOSOMES IN THE SPERMATOGENESIS OF THE HEMIPTERA HETEROPTERA. 



also other cliaracters of form, arrangement, and process change, some of which will 

 undoiihtedly be found to be of greater value than number in the analysis of descent. 

 McClung (1905) was the first to draw attention to arrangement of the chromosomes as 

 a high taxonomic character, thus seconding my idea (1901a, b) that there should be 

 a comparative phylogenetic study of the germ cells as a check and supplement to the 

 analysis of the phylogeny based upon somatic structures. The foundation of a rational 

 phylogenetic system upon cellular differences is as yet little mure than suggested, 

 because the comparative basis is so small and the phenomena so complex. Yet J 

 believe it should be attempted, and that it will be found to be entirely possible. 



Perhaps the best way of attacking the problem of the influences determining 

 chromosomal number, is by the analysis of the phenomena in those species where 

 thei'e are two normal numbers. 



In conclusion the position of the chromosomes in the equatorial plates of the 

 maturation mitoses of the Hemiptera may be sunnnarized. 



Those diplosomes that divide equationally in the first mitosis and reductionally 

 in the second are not central in the first spindle (except in Oncopeltiis), but are central 

 in the second spindle. 



Those diplosomes that divide first reductionally and second equationally are 

 always central in the first maturation spindle (except in the Reduviidte), and more or 

 less excentric in the second spindle (but central in the Reduviidse). 



It is therefore the rule that the positions of the diplosomes are reversed in the two 

 maturation spindles ; and that they are in the center of the chromosomal plate when 

 they are bivalent (except in the Reduviidoe). Consequently the position of the diplo- 

 somes is rather a criterion of their valence than a character of any taxonomic 

 importance. 



There is a tendency in most of the Hemiptera, when the autosomes are not very 

 numerous, for those of the first maturation spindle to be disposed in a circle around a 

 central one, while there is generally less regularity in the second maturation spindle. 

 Such positions would seem to be dependent upon the interaction of the number of 

 chromosomes and the mechanics of the cell division, and therefore to be of no particu- 

 lar taxonomic importance. 



