Manchester Memoirs, Vol, Ixii. (191 7), No. 2 



27 



co-efficients for the two- species given below are reproduced 

 from Agar's paper. 



Macrosiphum Antherinii. 



Parental 



117 parents 



and 



124 offspring. 



Grandparental 



54 grandparents 



and 



60 offspring. 



Hyalopterus Trirhodus. 



Parental 



60 parents 



and 



358 offspring. 



Grandparental: 



30 grandparents 



and 



291 offspring. 



Antenna 

 Frontal 



Breadth 

 Ratio 



0*482 ± o 046 



0-433 ±0049 

 0-235 ±0-057 



0-165 ±0*085 

 0*231 ± 0*082 



-0*002 ± 0*087 



C427 ± CO29 



0*439 ± 0*028 



0*177 ± °'°38 



0*321 ±0 035 

 0*230 ± 0037 



It will be seen that the two results show a. certain degree of 

 similarity, and, as Agar says, " afford] a certain presumption that 

 the co-efficients are due to the physiological relationship between 

 the grandparent, parent, and offspring, and not the accidental 

 results of extrinsic causes." 



Reasons are, however, offered for not accepting this conclu- 

 sion too readily; thus there is the possibility that the viviparous 

 nature of the animals and their short life history may have an 

 effect in the result; for the maternal nutrition may have in- 

 fluenced the size of the offspring, which was not able to attain 

 the normal one during the short life of the insect. Further, 

 the characters investigated were dimensional ones which are 

 easily influenced by environment. In this connection it must 

 also be noted that in Simocephalus exspinosus there was no 

 evidence of any diminution of the correlation coefficients as the 

 scale of ancestry was increased. 



The many variations which occur in Aphis avernas led Ewing 

 to study the inheritance of some of the characters of this insect. 

 The conditions in which the parthenogenetic females lived were 

 kept as far as possible constant. 



Pure lines were raised from these females and the mean for 

 the characters investigated was determined in each case and found 

 to be very constant. When extreme variants ojf a pure line were 

 selected to act as parents for further lines, the mean obtained 

 was the same as that of strain's which had not been specially 

 selected. Also the fraternal mean of a generation often exhibited 

 great fluctuatoins from, the mean of the strain, but these fluctua- 

 tions were not inherited. 



Finally in 191 5 a paper was published by McBride and Jack- 

 son on the inheritance of colour in la stick insect, Carausius 

 morosus. In this animal the colour patterns of the adults are 



