SPHINGID.E OF PERU. 81 



to the dead level of brown and withered stalks. The Interior at this time is bright 

 and only less hot than in the wet season, and in normal years is refreshed by periodic 

 and welcome showers. 



In my expeditions thither my endeavour was to test the country as far as possible at 

 different times of the year, and of the six which I made, that of August 1909, at the 

 end of an almost unprecedented period of drought for nearly four months, proved by 

 far the least remunerative. My list of dates shows that the months of November, 

 December, and January in the wet season are practically the only ones of which I had 

 no experience, and from what I have heard I gather that this is perhaps the best season 

 of the year from an entomologist's point of view. If one happens to be settled in 

 some headquarters beforehand, this is all well and good ; but with the lieavy rains, 

 derrumbes or land-slides on road and railway, paths knee-deep in mud and slippery 

 clay, and possibly a bridge or two washed away altogether, the four days' journey from 

 Lima to the Perene camp is often a sheer impossibility. From my last two expedi- 

 tions in 1910 it became evident that March and April were jiar excellence the months 

 for larvae, and though I got much in May, there wei'e very evident signs that I was 

 too late for many species. Quite the majority found were so nearly full-fed that it 

 was as much as I could do to keep up with the work of figuring them before they 

 changed. The captures of the day had generally to be pencilled in by lamp-light and 

 coloured next morning. 



For Lima and the west coast the seasonal changes are best shown by an examination 

 of the following table, which is constructed on the fairly complete records of my three 

 years' observations and the average taken. In the matter of abundance of specimens 

 the SphingidcE may be regarded as largely representative of other families of Lepidoptera. 

 It will be seen that the figures show a strange lack of sequence in increase and 

 diminution, making it hard to generalise, and still more difficult to state with reference 

 to any individual species how many broods it may have in the course of the year. 

 Its appearance as a moth or a caterpillar in so many months seems often to suggest 

 a continuous succession of broods. This is partly due no doubt to incomplete observa- 

 tion, but at least it becomes clear that from July to December is the poorest time for 

 larvse, and from July to February for imagines ; and also that March is undoubtedly 

 the best month in the year for Sphingidae in both the larval and imaginal condition. 

 It is a noteworthy fact that there is not one single month in the twelve when at least 

 a few species cannot be obtained in the one stage or the other. 



VOL. XX. — PART II, No, 2, — April, 1912, 



