34 



pression, others so deep that I could not remove the scales from them ; none of them 

 extending from base to outer angle of- wing. One would commence near the base, run 

 deep and terminate in a loop. Another would begin o|)posite to the middle of that one, 

 run be^'ond it and stop. Others formed plaits on the outer angle, all inclining to be 

 wavy, as if they had been subjected to longitudinal pressure. These foldings would ac- 

 count for the broken lines of the transverse corrugations. 



My next effort was to lay bare the nervures ; for this end I soaked the winglet in 

 water. It came out an elastic gelatinous mass. I tried hard to separate the upper and 

 under membranes but without success. I turned it over to examine the underside, and 

 found that the membranes had parted over some of the heavy nervures at the base of the 

 wing, disclosing their structure completely ; the end next the joint of one of them turned 

 upward, and I could see into the hollow tube as far as the bend would allow, the walls 

 appearing to be very thin ; inside they were smooth with a waxy look. A general sur- 

 vey of the exterior reminded me of an earth-worm severely contracted. A close inspec- 

 tion showed them to be segmented ; one end of the segment was prominently rounded, 

 sloping suddenly to the other end, entering the rounded end of the one next to it, and so 

 on along the portion of the nervure laid bare, without any perceptible diminution in its 

 size. If this structure is what is referred to in the statement, " the nervures themselves 

 are folded," then the term is badly chosen and very misleadina:. 



The condition of the costal edge of the winglet was now clearly displayed, but diffi- 

 cult to describe. If a piece of stiff twine is placed on the table, doubled back and forth 

 in as short bends as is possible to give it, kept in place and looked at from above, it will 

 give a good idea of the costa when it is viewed edgeways ; then cut the bends on the side 

 representing the centre of the wing, and it will have a strong resemblance to the upper 

 surface of the costal edge. The term, crimped, or crimpled, would well express its 

 condition. 



These views were obtained, and could only be obtained when the winglet was satu- 

 rated with moisture. We may find in this direction an almost unlimited field for obser- 

 vation. From what we know of insects, we should be warranted in expecting as' much 

 diversity to characterize this as any other department of their history. A general prin- 

 ciple may cover the whole, but with very considerable latitude in working out of the 

 details. 



The difference between the expanded and unexpanded wing is great in more ways than 

 size ; the wrinkled thickness of the nervures in the one, is in striking contrast to the smooth- 

 ness and fineness of the other. The amount of space between the rows of scales is very 

 suggestive ; in the unexpanded wing the roots of the scales are close together, each row 

 being situated on the crest of a wrinkle ; in the expanded wing the tips of one row of 

 scales just cover the roots of the row in front of it ; when the scales are removed, the 

 space between the rows of roots is a smooth dull coloured membrane. When contemplat- 

 ing this, I thought I saw an explanation of what I had observed when rearing Cecropias. 

 Occasionally I would get one of unusual size, but they were always thin in vestiture and 

 defective in colouring. Now if each specimen of a species has the same number of rows 

 of scales on their wings, which may be the case, and the membrane of one becomes un- 

 usually extended, without a corresponding increase in the size of its scales, we can easily 

 see what the consequences would be. v 



But the expansion of the scales themselves has to be accounted for. This takes 

 place at the same time as that of the wing. From whence comes the motive power 1 It 

 is known that the roots of the scales are bulbous, set in sockets, the bulb being under the 

 surface of the membrane ; and that the scales are double, open inside, but closed at the 

 outer end, forming a sac ; may not this opening extend through the stalk and root of the 

 scale, connecting with the space between the two membranes of the wing, and receive 

 from thence the impetus for its expansion 1 I tried to settle the question by various 

 observations but obtained nothing definite, yet this seems to be the direction in which 

 we have to look for the result produced. 



I bethought me of the Luna I had seen when it had just emerged from its cocoon, 

 which was pure white, resembling a bit of swan's down, so I examined the expanded wing 



