CKOSSING EUSCHISTUS VARlOLARIUS AND EUSCHISTUS SERVUS. 359 



compared with those, based on a study of the original specimens immediately 

 after the bugs were killed. The two sets of figures agree almost exactl}^, 

 differing only in the fact that in the original estimate we classed two speci- 

 mens as variolarius and two as servus which we now class as intermediates. 

 Such a variation in the classification is unavoidable because the intermediates 

 blend into the two extremes, and it is not always clear into which division a 

 specimen belongs. Possibly the printing of the photographs in the final plates 

 may be too light in some cases, and fail to bring out a faint spot that should 

 be classed as an intermediate and not servus, and this vrould cause some 

 slight variations from the above figures. If we consider separately the 

 ratio from each of the seven pairs of Fi hybrids, the results are as 

 follows : — 



F. 2 males from the seven pairs of Fj hybrids. 



1 



Spot ~^ 

 like 1^ 

 vario- ( 

 larius. J 



1st pair. 



Plate 29, 



Photos 



17-25. 



2ud pair. 



Plate 30, 

 Photos 

 26-32. 



3rd pair. 

 Plate 31, 



Photon 

 33-41. 



4th pair. 

 Plate 32, 



Photos 

 42-48. 



5th pair. 



Plate 28, 

 Photos 

 16-16. 



6th pair. 



Plate 32, 

 Photos 

 49-50. 







7th pair. 



Plate 33, 

 Photos 

 51-57. 



Totals. 



1 



3 



8 



4 



1 



2 



19 



Inter- 1 

 mediates, j 



17 



12 



27 



18 



2 



6 



IG 



97 



No spot 1 

 like I 



25 



15 



13 



5 



1 



1 



14 



74 



servus. \ 



















Of these seven pairs of F^ hybrids only the 5th pair shows a simple Mende- 

 lian ratio (1 like variolaiius, 2 intermediate, and 1 like servus), a ratio which 

 evidently would not have been maintained if a large number of offspring 

 had been secured, for no such ratio is shown by the Fg hybrids from any 

 of the remaining six pairs, all of which have a larger number of offspring. 

 If one is willing to readjust these ratios by filling the gaps with the 

 nymphs that died and those that failed to hatch, it would be possible 

 of course to create any ratio required, but the assumptions cannot be the 

 same for the seven pairs, and this fact would weaken the value of such an 

 adjustment. 



y\^ e may further search for a Mendelian ratio (half pure and half hetero- 

 zygous) in the cross between the F^ hybrid $ and the pure variolarius (^ . 

 From this cross we raised 18 males (photos 62-66). 12 of these 18 males 



