86 THE entomologist's record. 



ilione and Metlmua confusa already examined, are the models. In the 

 Dhmorphia the transparent portion shows the scales reduced in size 

 only, the nmnbers not being reduced nor the shape altered. In the 

 Hypsids, all species of the genus A7ithi»iii/za have the shape, numbers, 

 and disposition of the scales the same, and transparency is effected by 

 the scales themselves becoming transparent. In remarkable contrast 

 with this, Hyelosia tiresia, also a Hypsid, shows the scales greatly reduced 

 in number as well as being transparent, and also set up on edge to a 

 certain extent. The last species to be examined by Professor Poulton 

 were certain of the Castniidac. Several transparent species were 

 examined, and all had the same arrangement. The scales are not 

 reduced in size, and but little in number. The shape is simplified and 

 the pigment is lost. The scales are also set steeply up on edge. 



Several papers have been written on the pigment in lepidopterous 

 scales, and some of these indirectly touch on transparency. Such are 

 Mr. A. G. Mayer's " Development of the wing-scales and their pig- 

 ment in butterflies and moths " (see Mmeum of Comp. Zool. at 

 Harvard Coll., Camb., U.S.A.), and Dr. T. A. Chapman's " Notes on 

 Mr. Mayer's paper" {Knt. Bee, ix., pp. 78-79). In a paper, however, 

 by Mr. E. W. Wake Bowell {Ent. Rec, ix., pp. 271-^73), on the 

 " Scales of the European Zyg^nides," the shape and disposition of the 

 scales are discussed, and the question of transparency and its develop- 

 ment is unconsciously dealt with. I say unconsciously because of the 

 statement that "As usual, the scales of the lower wings are less 

 highly specialized than those of the upper." This, as I shall show 

 later on, is the converse of what has and is taking place. Mr. Bowell 

 finds that the trifid, bifid, or quadrifid scales of the forewing are more 

 specialized than the strap-shaped smooth scales of the hindwing, 

 which practically means that the forewing once had simple scales and 

 has developed elaborate scales. One need only look at a very 

 completely transparent hindwing covered with minute hairs to see 

 the impossibility of such an argument. When I began to write this 

 paper I had a vague recollection of having seen a paper on the scales 

 of the bee-hawk moths, but a perusal of the most likely literature has 

 failed to reveal it, so at the risk of recapitulation, I will give a few 

 details of these interesting Sphingids. 



In Hemaru fuciformis the scales in the transparent portion are fairly 

 closely packed together on emergence from the pupa but have lost 

 some of their pigment, and some are semi-transparent, but as the 

 sockets are very weak and the scales themselves rather elevated, they 

 quickly wear off and natural selection will probably not work apy 

 further to make the now fugitive scales themselves transparent. In 

 all the species of He ma) is 1 have examined the same means is adopted, 

 except that in some species notably in H. aifinis, from North America, 

 the scales are not nearly so elevated and, therefore, not so likely to 

 brush off, and one might expect that the scales themselves might be 

 more transparent, yet this does not appear to be so. In Hemaris 

 radians, from Japan, the scales in the transparent area are apparently 

 considerably more transparent, at any rate in the interspaces between 

 the nervures, but I have been unable to examine a well-scaled 

 specimen. In Cephonodes Jii/las the transparency of the scales has 

 developed to a considerable extent, but I have only been able to 

 examine specimens with a few scattered scales still left on the 

 membrane. {To be concluded.) 



