260 Transactions of the 



certain whether one or two membranes were under view, thou^^h 

 evidently some were torn apart ; nor was it possible to find with 

 the necessary certainty the corresponding portion of the scale on 

 the opposite glass. Perhaps if tried using only a few scales on a 

 thinly-varnished shde, employing a solution of wax in benzole or 

 chloroform, and covering them with a similarly-prepared stout 

 cover ; when separated by force, the scales might be split. This is 

 only a suggestion arising from the previous experiment. 



The examinations have been most fruitful on highly-dried or 

 old specimens from hot climates, when anatomized scales were 

 sought. The torrefying alluded to in the previous part has been 

 managed in the following way: — A clean cover had some scales 

 spread out on the centre and examined to see if it contained both 

 the simple and wavy kinds ; then covered by a rather smaller thin 

 cover, and both carefully placed on a flat thin piece of iron plate 

 (the u'on frame of the large linen-covered buttons answered well). 

 This was held by the forceps at one side, and gradually introduced 

 at its edge into the flame of a spirit-lamp. In a short time the 

 scales were seen to suddenly change colour, and become almost lost 

 to the naked eye. At this moment the support was withdrawn 

 from the flame, but kept at a greater distance from the source 

 of heat, to thoroughly dry the scales. Without the precaution of 

 covering the scales by a second thin glass, they were generally found 

 so contorted as to be useless. Much care is required, or they become 

 fused. The ribs sometimes separated unequally, the interspaces 

 often giving rise to the optical efi'ect of the " notes " ! ! {vide Fig. 18) ; 

 others shrivelled in various directions, and produced, with some 

 illuminations, most charming chromatic efibcts; others appeared 

 melted on to the surface, so that a needle could be made to divide 

 them or traverse them in any direction, without displacing the 

 scale. In some, nothing could exceed the beauty of the illusory 

 beaded appearance, both longitudinal and oblique. An attempt has 

 been made to figure the efi'ect (Figs. 13 to 19) ; and in others the 

 transparent membrane was shown with a portion of the framework 

 attached. On bruising a thoroughly-roasted scale, there ivas no 

 evidence whatever of any heads heing in or sejyardble from the 

 scales. The particles of pigment that became detached in every 

 case, most likely from the pigmentary deposit, were very irregular 

 in size and shape [vide Fig. 8, from an untorrefied scale). 



As regards the nature of the pigment, so called, it is feared 

 the views now ofiered may not quite correspond with the usually- 

 received opinions. A careful examination with reagents is needed. 

 I have not been able to separate it in quantity sufficient for this 

 purpose, though my friend. Dr. Southby, writes me from Italy, 

 where the Humming-bird Sphinx Moth {Macroglossa stellatarum) 

 was plentiful, that he had been enabled to displace it from the sur- 



