192 Edmund B. Wilson 



interpretation be correct, the origin of an unpaired chromosome 

 in certain individuals of this genus has been owing to the same 

 cause that has produced the supernumeraries. Since both condi- 

 tions coexist in the same species, along with that which may be 

 regarded as the original type (22 chromosomes) it may be conclu- 

 ded that Metapodius is now in a period of transition from the sec- 

 ond to the third of the types distinguished in my last study. It 

 seems quite possible that other species of coreids that now have 

 constantly an unpaired idiochromosome may have passed through 

 a similar condition, though in all of them thus far examined both 

 the small idiochromosome and the supernumeraries have dis- 

 appeared. In Metapodius, accordingly, the supernumeraries 

 may be regarded as on the road to disappearance. That such is 

 the case is rendered probable by the fact that their number does 

 not pass a certain limit, and is rarely more than four. The very 

 small chromosomes of this kind, so often observed, are perhaps 

 degenerating, or even vestigial in character. But aside from this, 

 attention has already been called to the probability that one or 

 more of the supernumeraries may be lost during the growth-period 

 (p.i86); and while this is not certain, it may well be that both 

 methods are operative in their disappearance. 



The foregoing interpretation of the supernumeraries enables 

 us to understand why variations in their number are not accom- 

 panied by corresponding morphological differences in the soma- 

 tic characters; for they are but duplicates of a chromosome already 

 present and hence introduce no new qualitative factor. It can 

 hardly be doubted that some kind of quantitative difference must 

 exist between individuals that show different numbers, but none 



has been more frequently met with. Another objection might be based on the different relations that 

 occur in Syromastes. In this form (see Wilson '09) the passage of both idiochromosomes to one pole 

 without separation is a normal and constant feature of the second division, yet no supernumeraries 

 appear in any of the individuals, and it is probable that the female groups contain two pairs of idio- 

 chromosomes like the single pair that appears in the male. We have no data for a conjecture as to how 

 such a condition can have arisen; but evidently the small idiochromosome does not in this case become 

 an erratic supernumerary but retains a definite adjustment to the other chromosomes. Still, I do not 

 consider this an obstacle to my interpretation of Metapodius, for it is'now evident that the history of the 

 idiochromosomes in general has differed widely in different species and families, even among the Hemip- 

 tera. We have thus far only made a beginning in their comparative study. [See Addendum, p. 200.] 



