516 Mr. G. A. K. Marshall on 



complete illustration hitherto known of the power of 

 mimicry to attract forms of all kinds irrespectiv^e of 

 affinity. It is of the deepest interest to observe that the 

 association is almost entirely Mtillerian (synaposematic). 

 The following passages are quoted from letters written by 

 Mr. Marshall about the time when the first consignments 

 were sent ; but the group became very much larger when 

 the additions of later consignments were included. — 

 E. B. P.] 



Salishiivy, Jan. 8, 1899. — The two most prominent types 

 (■f colour among mimetic Coleoptera are what I call the 

 Lycoid and Mutilloid types. Of the former I have put 

 aside for you a series comprising fifteen species, including 

 .several species of Lycus, three species of Longicorns, a 

 ileduviid bug, a fly, a wasp, an Arctiid moth, a Mylahris, a 

 IFypcracantha, etc., and I shall be able to add more in all 

 probability. 



The Lycoid Arctiid moth is a day-flyer, but the deceptive 

 icsemblance is not good on the wing, being best shown 

 when the moth sits, as is its wont, on the ends of grass- 

 stems, etc., after the manner of a Lycus; it is perhaps 

 even more like rrionoccrus dimicUatvs, a Lycoid unpalat- 

 able Malacoderm which has a similar habit. 



Salisbury, Fch. 12, 1899. — The six species of Lycus, the 

 Prionocerus, Diacantha, Zonitis, Mylahris, and Elctica, I 

 proved by experiment to be distasteful to baboons and a 

 kestrel. The Zyga^nid I presume to be so likewise, as it 

 emits a strong smell ; the Tckphorus will also probably 

 prove to be unpalatable. The four species of Avasps have 

 all got very effective stings ; thus the only unprotected 

 insects are the fly, which is an adiuirable mimic of one of 

 the wasps, and the three Longicorns, though I am not quite 

 certain about Philagcthcs. As to the Peduvius I do not 

 know what to say, there are certainly some very remarkable 

 cases of mimicry in this family. ♦ # * * 



In flight the Zygaiuid \_Nrnrosymfloca, Fig. 52] is aided by 

 its very brilliant hiiid-wings, and the H3-menoptera have a 

 fli<dit very difterent from, and far swifter than, that of 

 J/ycus. 



[A complete list of the species arranged in their respec- 

 tive families is given below. Large as the group is it 

 could certainly be made much larger, especially if the 

 whole of South Africa were put under contribution. Thus 

 an obvious addition to the Longicorn mimics is Dyenmo- 



