No. 521] 



NOTES AND LITERATURE 



311 



their values are subsequently used in the discussion on the 

 assumption that the material is really a random sample of the 

 general population. The discussion of the biological import of 

 the constants has no significance if this assumption is not true. 

 This being so, it is not feasible to turn around and say that the 

 values of the constants prove the randomness of the material. 



On the basis of this rather dubious material the discussion of 

 a very interesting question of inheritance is undertaken. The 

 characters dealt with are the same as those of the former paper. 

 The first point brought out is that there is a slight but definite 

 differentiation between the "single nest" and the "general pop- 

 ulation" queens. The authors state that they are not clear as 

 to what is the cause of the differentiation. The main point of 

 the paper turns on the relative variability of the two groups. 

 Taking first the standard deviations (the coefficients of variation 

 show the same thing), it appears that the "single nest" queens 

 are roughly only just about half as variable as the "general 

 population" queens. A long discussion follows regarding the 

 question of whether the amount of this reduction in variability 

 in the single nest is what would be expected on the basis of the 

 law of ancestral inheritance. The general outcome is that it is 

 not ! There are, however, so many biological factors about which 

 the authors avowedly have no data at all, concerned in the pro- 

 duction and interpretation of the observed results, to say noth- 

 ing of the general difficulty about the randomness of the 

 "general population" sample, that the discussion fails to be very 

 convincing in any direction. This particular case well illus- 

 trates a tendency which seems likely to do a good deal of harm 

 to the biometrical cause, so far, at least, as biologists are con- 

 cerned. The whole discussion of inheritance in wasps in this 

 paper rests on a series of premises and assumptions regarding 

 wasp biology, which are made without any attempt whatever first 

 to learn by direct investigation the actual biological facts. This 

 method of developing a long and involved theoretical argument, 

 with very far-reaching ultimate conclusions, upon an exceedinglv 

 slender basis of facts, gives such discussions of heredity a highly 

 academic, not to say Pickwickian, flavor. In general one can 

 not help feeling that if one becomes seized of a desire to know 

 how characters are inherited in wasps the direct and straight- 

 forward way to set about easing his intellectual pangs is to breed 

 wasps under controlled experimental conditions and observe the 

 results. The technical difficulties of dealing with this particular 

 material in this way would doubtless lie considerable, but prob- 



