438 Mr. G. A. K. Marshall on 



Anomctlijms, etc. These insects appear with a rush as 

 soon as the early rains have saturated the ground, but 

 should a dry spell supervene, they disappear as rapidly as 

 they came, only to emerge again on the recurrence of a 

 good rain. The case of the dung-beetles has always 

 puzzled me, for here we have a large family of powerful 

 and apparently hardy beetles, which have a constant 

 supply of food all the year round, and yet they are unable 

 to stand out the winter in the imago state, although a 

 delicate butterfly can do so. Jn fact, the Gopridfe are 

 quite as dependent on moisture as the large Carahidm, and 

 are only to be seen at work from November to March, 

 retiring even then during the dry spells." 



The interesting effects of dryness described above cer- 

 tainly cannot be produced iu our damp winters, and it is 

 difficult to believe that our cold can be the cause of the 

 retirement of Goccincllidie, etc., when species of insects 

 closely allied to those of England can endure to be frozen 

 stiff and brittle iu a temperature of 50 degrees below zero 

 (F.) in a Manitoban winter. 



Another letter from Mr. Marshall, received about the 

 same time, contains a different comment upon the inter- 

 pretation suggested by the present writer iu 1898 and 

 here amplified. 



" Salisharjf, Jan. 8, 1899. — I can fully perceive that 

 any arguments that may be brought forward in support 

 of the contention that Precis sesarmts (natalensis form) 

 is an e.xample of incipient mimicry are equally applicable 

 to the suggestion of incipient warning coloration, and 

 for the present it must remain a matter of opinion as to 

 which is the correct explanation, though the alertness of 

 the insect and its undoubted palatability, so far as lizards 

 are concerned, seems to lend more support to the former 

 view to my mind." [See also Ann. and Mag. Nat. Hist., 

 ser. 7, vol'ii, July 1898, p. 35.] 



It has been here shown that there are important elements 

 in the under-side coloration of the wet phases of Precis 

 scsamus and P. antilopc which cannot be explained as 

 mimicry, Batesian or Mlillerian (see pp. 425-8), while the 

 entire appearance of the under surface of P. ctrcliesia form 

 pclnsqis can only be interpreted as a warning character 

 (pp. 428-431). 



The conspicuous appearance of the under-sides of these 

 foiins is doubtless chiefly adapted to render them con- 



