EXPKEIMENTS UPON ACR^A ENCEDON. 399 



appear are transitional between commLxta and infiiscata on the one side, and 

 rommiMa and li/cia on the other, rather than directly between the two chief 

 forms. Hence, by selecting the examples, a fair transition from infiiscata to 

 hjcia may be constructed by way of commixta. This latter form is clearly 

 hereditary. Thus the female parent of Family 7 is commixta, and all 3 of 

 her non-lycia offspring are commixta. Particularly interesting in this respect 

 is Family 5, of which the female parent possesses the fulvous bar but not the 

 Avhite hind wing of commixta. About half of the offspring, namely 13 out of 

 28, resemble the mother in this respect *, while 2 of them have also the white 

 hind wing. Other evidence of the hereditary transmission of this combination 

 o£ characters will be found under Family 2 (pp. 407-8). 



Families of loliicli the female parent only is known. 



When the only known parent is a female infnscata (or commixta) the off- 

 spring show, in all four families, a mixture of infiiscata (or commixta) and 

 lycia with approximate equality three times. In the single exception, 

 Family 7j the numbers are small. The inference is that one parent was 

 recessive and the other heterozygote. 



When the only known parent is a female lycia, the offspring are all lycia 

 twice and mixed li/cia and infiiscata ioiw times, with equality thrice (including 

 the small Family 9) and irregularity once (Family 8). 



In the absence of selective breeding, for which there is insufficient evidence, 

 the great numerical superiority of lycia would result in the majority of the 

 pairings being between males and females of this form, or between lycia 

 and infiiscata, the latter being far more commonly heterozygote than pure 

 dominant. We can thus, on the hypothesis that lyciais recessive, understand 

 why the families bred from a female of this form were either all lycia or mixed 

 lycia and infiiscata, but, owing to the relative rarity of the pure dominant, 

 never, in the author's experience, entirely made up of heterozygotes bearing 

 the appearance of the dominant (infuscata). It is unfortunate that the 

 infiiscata in the families with equal numbers of the two forms never happen 

 to have been bred from, so that their heterozygote constitution could be 

 tested. 



Families of lohich both parents are known. 



Both male and female are of the form lycia in 8 families ; both are of the 

 form infuscata in only 1, namely Family 5. The 8 former produced only 

 lycia offspring ; the latter only infuscata. 



* The numbers suggest that, as regards this character, the parents were recessive and 

 heterozygote respectively. If this be so, the female, belonging to the rarer form, was 

 probably heterozygote, while infuscata, although dominant in relation to It/cia, is recessive 

 to the form with a fulvous bar. 



