320 Dr. Frederick A. Dixey on 



latter bar is indicated, the chestnut-coloured streak takes 

 no part in its formation ; in P, le'ptalina, however, the 

 same chestnut-coloured streak is better pronounced along 

 the hind margin than in P. loandosia, and is prolonged 

 for a short distance along the transverse bar (fig. 2, o). 

 In this way a considerable resemblance is bi'ought about 

 between P. leptalina and some of the well-known trans- 

 parent neotropical Danainee, which are presumably 

 protected, such as Napeogenes inachia (fig. 4). It is 

 worth noting that the bar o, which belongs to the hind- 

 wing in P. pandosia and P. leptalina, appears to 

 represent the dark inner margin of the forewing in the 

 model ; and the presence upon its distal portion in P. 

 lep)talina of a slight prolongation of the chestnut marginal 

 streak may pei^haps stand for the turning inwards of the 

 corresponding chestnut marginal band in N. inachia 

 along the costal border (figs. 2 and 4). 



4. The " numata " line. This series needs here no 

 more than a mention, as it has been already discussed 

 and illustrated in the paper above referred to.* 



5. The " tarracina" line. This is an oflfshoot of the 

 last, or numata line, diverging in the neighbourhood of 

 Mylothris malenha $ , or perhaps somewhat further back 

 towards the original Pieris. The remarkable form M. 

 alethina ^ shows a near approach to the pattern of 

 Tithorea tarracina and other associated species; its own 

 aspect being probably derived from a type like M. lorena $ 

 or AI. malenka ^ , in which the yellow of P. deivop)hile $ 

 has persisted, and the base of the forewings has become 

 overspread with black. t 



General Considerations. There is, therefore, in each one 

 of these cases, a continuous line to be traced ; starting 

 in every instance from the same ordinary Pierine form, 

 and passing through a graduated series of closely-allied 



*' Vide supra, p. 319, note. 



t Three at least of the above five lines of development, viz., 

 the agna, the numata, and the inachkt lines, can be paralleled from 

 the genus Bismorphia and its allies, and it is interesting to see in 

 the latter case how the same ultimate result is brought about by 

 somewhat different means. But I have not attempted to include 

 species of Dismorphia within the above series, because the afiinity 

 of this genus with Pieris and Mylothris is not close, and its own 

 course of mimetic development, so far as phylogeny is concerned, 

 must be regarded as completely independent. 



