4ii 



SNODGRASS 



melanocera on Charles. The characters of the races of each species 

 as given above do not throw any light on the question as to which is 

 ancestral in each case. Since the Iguana Cove and Indefatigable races 

 of S. melanocera differ from the others in having a pale abdominal 

 band, it is probable that they are secondary ; and since the Indefatig- 

 able race has the unique character of an entirely black prozona, it has 

 probably been derived from the Iguana Cove race. Otherwise we 

 receive no help on this point fi-om a study of the specimens themselves. 

 Any form, as far as we can see, could be the direct descendants of the 

 ancestors of the others in the case of either species. 



If Charles and Chatham Islands are, as they appear to be, the oldest 

 habitable areas of the archipelago, then it is most probable that they 

 were the first islands inhabited by representatives of Schistocerca . 

 Adopting this view, it is very easy to explain the status of the genus 

 at present on the archipelago. We can suppose that the original form 

 became modified on Chatham to S. Uterosa and on Charles to S. 

 jnelanocera . From Chatham, we can suppose, individuals migrated 

 to Tower and to Hood — each island being more accessible from Chat- 

 ham than from Charles — and on these islands became the races peculiar 

 to them. In the same way representatives from Charles could have 

 gone to Harrington, Tagus Cove and Iguana Cove, and from the 

 latter place to Indefatigable. To explain the Duncan race we have 

 only to imagine that individuals from both Chatham and Charles were 

 landed upon Duncan, thei-e producing its hybrid literosa-tnelanocera 

 race. Finally, individuals from Duncan could have gone to Abing- 

 don and Bindloe Islands and there produced the Abingdon-Bindloe race. 

 This relationship is diagrammatically shown on Plate xxvi. 



It is evident that we might retain the same lines of relationship, but 

 turn the direction of migration the other way. That is, we might sup- 

 pose that the Abingdon-Bindloe race is the ancestral one, that from it 

 was produced the variable race on Duncan, and that individuals repre- 

 senting one extreme of this race went to Chatham or Hood and indi- 

 viduals of the other extreme to Charles. From these islands the other 

 varieties could radiate as before. The chief objections to this view are 

 first that Abingdon and Bindloe are not very old islands, and second 

 that migrants would have both the prevailing winds and currents 

 against them. If the archipelago were inverted in position so that the 

 northernmost islands were farthest south and vice versa^ and if the then 

 southern islands appeared to be the oldest, we could easily regard the 

 Abingdon-Bindloe race as the unmodified original species, the Duncan 

 race as a variable form produced from the first, and the others as being 



