600 PEOF. E. B. POULTOJSr : NATTJEAL SELECTION 



has reached a very different level in the two species. In this, 

 as in all other instances, the transparent patch at the apex of the 

 left fore wing was selected for study and for photomicrography. 

 Dr. Gustav Mann, of the Oxford University Physiological 

 Laboratory, kindly assisted me in the difficult details of this 

 latter work, and such success as I have been able to attain I owe 

 mainly to him. Eig. 2 shows the structure of the transparent 

 patch in Methona covfusa ; fig. 3 its structure in Thyridia 

 psidii. In both the broad scales are reduced to small bifid 

 structures, of a distinctly different form in the two species ; 

 while the narrow scales are reduced to long extremely fine hairs 

 iu the Methona, and to much broader ones in the Thyridia. 

 In both the transition into the scales of the opaque part of the 

 wing is well seen on the upper part of the figures. The 

 comparison between the almost identical appearance of the two 

 insects as shown in fig. 1, and their real differences which are 

 seen to be very marked in figs. 2 and 3, afi:'ords another good 

 example of the principle which has been already so abundantly 

 illustrated — that when a close superficial resemblance has 

 been attained no further similarity iu the details of structure 

 is produced. This is apparent enough even in these two genera 

 belonging to the same sub-family. It is of course illustrated in 

 a far more striking manner when the affinity is much ujore 

 remote. 



We now pass to the Danaince which fall into the group. 

 The BanaincB supply it with two species, both belonging to 

 the genus Ituna, shown in PI. 43. fig. 4. The Ithomiince were 

 formerly classed with the Danaince, and in any case form the 

 sub-family which stands next to them. But fig. 5 shows that 

 transparency is attained in an entirely diflferent manner, the 

 scales being neither greatly reduced in size nor much altered 

 in appearance. It is the great reduction in numbers which 

 is here the cause of the transparency. The scales retain their 

 dark pigment and produce the eft'ect of a grey dusting over 

 the transparent areas. The transition into the opaque black 

 part of the wing is well seen on the upper part of the figure. 



Einally, there is a single species belonging to the distant 

 family Papilionidee and sub-family JPierince — Dismorpliia orise. 

 This species is shown in PL 44. fig. 1 together with Methona 

 confusa. The upper two figures of the Methona and of the 

 Dismorphia show a male and female of the cype form of each 



