TEPHROSIA CREPUSCULARIA. 29 



white cloud upon which the sun is shining." And, again, 

 "Direct sunlight is not to he employed, and a very strong light 

 of any kind is hurtful to the eyes." Under this advice I set to 

 work on a brilliant July day, my sitting lasting six hours, to see 

 if there could be any difference between the wing-scales of ere- 

 jmscidaria (bistortata) and biundularia. The lenses used were 

 chiefly | in. for individual examination, and 1£ in. for viewing 

 "the field." And how can we best examine, microscopically, the 

 scales of an insect ? Take a glass slip, 3 in. by 1 in. ; breathe 

 upon it, and press gently but firmly an upper and lower wing of 

 the set specimen on the part of the slip breathed upon ; lift the 

 wings (the moth is as good as ever) and you will just see a little 

 patch of dust on the slip. This is a facsimile of the wings, con- 

 sisting of scores upon scores of scales. Place the slip on the 

 stage of the microscope, and adjust the instrument to suit your 

 sight. Then take pencil and paper and sketch what you sec, for 

 memory cannot be trusted in these matters. Never did trans- 

 formation scene equal the sight you behold ! In the sun — for we 

 ivill do the thing we ought not to do — and with a 1| in. lens, the 

 scales glisten like burnished silver. We are looking at "a 

 field " of them. They are all perfectly distinct in outline, but 

 the sight is too much for mortal eye, so we choose the light from 

 " a white cloud," and the silver becomes mother-of-pearl, with 

 all sorts of prismatic colours. Still deeper we go with the |- in. 

 lens, and the scales are seen to have a central vein or nervule, 

 with close, parallel, but dotted lines throughout their entire 

 length. With this lens I made out the dots to be little depres- 

 sions with raised centres — but here I speak from memory. 



The shape of the predominant wing -scales* on crepuscularia 

 {bistortata) and biundularia is exactly like that of the blade of a 

 scull broken off near the top of the shaft. The end, or termina- 

 tion, however, is prominently toothed. These teeth may be two 

 in number — in which case they are of equal length — three, four, 

 or five, and then the central one, two, or three of them are 

 longer than the others. A rarer form is the well-known " battle- 

 dore " scale, and a still rarer one has the outline club-shaped. 

 The longer fringe-scales (cilia) are, perhaps, the most beautiful of 

 all. Some are shaped like a blade of grass, but the majority are 

 deeply serrated or toothed at the extremities — a sketch before me 

 shows four, another six, serrations — in fact, the cilia are perfect 

 miniatures of the petals of a pink. The discovery of a fresh 

 scale, of course, meant testing the whole of the specimens over 

 again until it was found in all. This multiplied my labours, 

 which did not end until the 30th of October. 



Summarising, then, I found all the Tephrosias, including 

 consonaria, luridata (cxtersaria), and punctulata, had scales of the 



* Their length = breadth x 3. 



