group of Butterflies, Potaro District, British Guiana. 415 



allied undescribed species. There must be much specula- 

 tion as to whether M. mncme or M. crameri commenced to 

 draw the many other species to them in coloration and 

 pattern. Mdinssa mneme is a strong variant in both fore- 

 and hind- wings while M. crameri is very constant above 

 and only as a very rare aberration is a form found with 

 the black of the hind-wing divided by the ground colour 

 so as to form a band. On the under-side, however, there 

 is considerably more variation. The latter species, owing 

 to its comparative constancy, must be looked upon as 

 older than M. mneme, a very variable and apparently 

 unstable species. It is however certain that if Af. crameri 

 was first in the field, 31. mneme must have entered soon 

 after, for many of the associated species of other genera 

 follow M. mneme to a greater extent than M. crameri. 



In fact M. mneme must have been far more potent than 

 M. o^ameri, and the strong variability must have been a 

 great factor in drawing so many different species to the 

 association. Of M. egina there is every reason to suppose 

 that it became a fixed and well-defined species early in 

 the history of the group, for we find only one other species 

 closely following it, and that also is a usually very constant 

 species, Heliconius silvana. It should here be mentioned 

 however that two specimens of Heliconius have been 

 caught, one in March 1905, the other without date, which 

 appear to be aberrations of H. silvana with a distinct 

 transverse black band to the hind-wing. These un- 

 doubtedly point to a not very distant genetic relationship 

 with Heliconius numata. 



Unquestionably the closest " pairs " are the Melinsea 

 with Heliconius species. Mcchanitis follows them very 

 closely with Lycorea also. While Etieidcs, Ceratinia and 

 StalacJitis, in the order named, diverge more and more 

 from the protected pattern. The identical pattern and 

 colour in some of the forms of Heliconius numata to M. 

 mneme is remarkable, as in the Potaro district the Heliconius 

 is apparently never abundant, rarely even really common. 

 I have only 32 specimens, and this represents the whole 

 take. The series is most remarkable for the very ex- 

 tensive variation, some having a narrowly barred hind- 

 wing, others having almost the whole of the hind-wing 

 black except for the costal portion. The Lycorcas are 

 certainly more abundant, while the Mechanitis species, 

 both pannifera and polymnia, occur in large numbers. Of 



