INTRODUCTION. x 



more decidedly red-brown. Another close connection exists between 

 yellow and the purplish-brown of Lasiocampa quercifolia, as I have a 

 yellow variety of this species. 



I would also add, in connection with the change from red to 

 yellow, a remarkable circumstance, first made known to me some 

 years ago by Mr. Coverdale, that is worthy of notice, viz., that although 

 several reagents will turn the red of the various Zygcenidce and their 

 allies, yellow, a careful washing with water, and even long continued 

 exposure to air, will in some instances restore the colour. This, of 

 course, only occurred in a very limited number of reds, the greater 

 number being altered quite permanently. I was interested to see that 

 Mr. Coste had lately noticed the same phenomenon. 



There is another co lour which the study of variation explains very 

 considerably. I refer to grey. Sometimes grey is the result of a pure black 

 and white surface absorption and reflection respectively, and is then 

 probably produced without any pigment. But there is another form of it 

 sometimes called stone-grey, ochreous-grey etc., which has a somewhat 

 similar superficial appearance to pure grey (formed from black and 

 white), with the addition of the slightest suspicion of ochreous in its 

 composition. That this tint is always in reality a yellow, or has a 

 yellow base, is I think obvious, if we study the variation. This par- 

 ticular phase of coloration is most marked in certain species such as, 

 Tceniocampa gracilis, T. cnida, Tapinostola hellmanni, Orrhodia erythroce- 

 phala, Leucania albipuncta v. grisea and L. lithargyria v. argyritis, 

 Grammesia trigammica, Mamestra sordida, Tceniocampa populeti (an 

 extreme form), Xylophasia polyodon, Dyschorista ypsilon, Orthosia 

 lota (an extreme form), Cosmia trapezina &c., and in all these, we 

 find, that these stone-coloured forms have almost always distinctly 

 red, and very frequently both red and yellow varieties, whilst the 

 transition in some of the species just mentioned is continued 

 onwards to deep red-brown and black. I consider the grey of 

 Lithosia griseola and its allies identical in character, although the 

 continuation to red does not take place, and the probability is, that if 

 the paler forms of the before mentioned species are now in a state of 

 development, tending in the direction of colour retrogression towards 

 white, the latter must be in a state of development, probably in the 

 direction of colour advancement. Another experiment that strikes me 

 with relation to this colour was an accidental one I made on Acidalia 

 ochrata. This species is normally of a deep orange almost brown 

 coloration. Having captured a considerable number, which I could 

 not set immediately, I pinned them into a zinc box in which a few 

 drops of liquid ammonia had been poured. Opening the box shortly 

 afterwards, the whole of the specimens were changed to a pale grey colo- 

 ration, with the slightest trace of ochreous very much like normal A cidalia 

 inornata. Exposure to the air, however, restored the colour entirely, 

 and the specimens were in a short time as richly coloured as ever. 

 Here ammonia effected artificially a retrogressive change, which is 

 very suggestive when we consider that A. aversata and its reddish 

 form exhibit in nature, the two colorations which ochrata exhibits when 

 treated with ammonia artificially and in its natural condition. It seems 

 certainly a fair assumption that the colour of A. ochrata has developed 

 from such a pale shade, and that the red var. of aversata is an attempt 

 in the latter species to follow the example of ochrata. The intimate 



