150 Edyniind B. Wilson 



The material of terminalis is from New Jersey, Pennsylvania, 

 Ohio, North Carolina, South Carolina and Georgia; that of femor- 

 atus from the three states last named; that of granulosus from 

 Arizona. The variation of number is independent of locality, and 

 individuals of the same species show^ing different numbers were 

 often taken side by side on the same food plants. It is equally 

 independent of sex, as the table at once shows. I am unable to 

 find any constant correlation between the number of chromosomes 

 and any other visible structural characters of the adult animals. 



Such an astonishing range of variation in the chromosome num- 

 ber in the same species seems at first sight to present a condition 

 of chaotic confusion. But, as I shall endeavor to show, the first 

 impression thus created disappears upon more critical examination. 

 Detailed study of the facts proves that the variation is not indis- 

 criminate but affects only a particular class of small chromo- 

 somes that are distinguishable from the ordinary ones both by 

 size and by certain very definite pecuharities of behavior. These 

 chromosomes are absent in all of Montgomery's material; in my 

 own they are sometimes present, sometimes absent, the total num- 

 ber varying accordingly. The chromosomes in question are the 

 ones which in earlier papers I have called the "supernumeraries."^ 

 In behavior they show an unmistakable similarity to the idiochro- 

 mosomes; and for reasons given beyond I believe them to be noth- 

 ing other than additional small idiochromosomes, the presence of 

 which has resulted from irregularities of distribution of the idio- 

 chromosomes in preceding generations. The relations seen in 

 Montgomery's material form the converse case, the small idio- 

 chromosome having disappeared or dropped out. I shall try to 

 show that both cases are probably due to the same initial cause. 



^ Wilson '07a, '07b. I first discovered this phenomenon in the pentatomid species Banasa calva 

 ('05b) describing the single supernumerary as a "heterotropic chromosome." Later ('07a) a single 

 supernumerary was found in certain individuals of Metapodius terminalis, and other numerical varia- 

 tions in this species and in femoratus and granulosus were briefly recorded; but at that time I did not 

 yet fully understand the facts. Banasa calva is the only form oustide the genus Metapodius, in a totalof 

 more than seventy species of Hemiptera I have examined, in which supernumerary chromosomes have 

 been found. Miss Stevens ('08b) has recently found in the coleopteran genus Diabrotica a condition 

 that is in some respects analogous to that seen in Metapodius. 



