BATTLEDORE SCALES OF LYCiENIDiE. 129 



upon them as flat surfaces, it would be so with many ; but 

 they are more or less globose or cylindrical, and have 

 manifest rotundity. I prefer the name of plumules, given 

 to their congeners among the Pieridse. 



They are most beautiftd microscopic objects, and interest- 

 ing in a physiological sense, displaying how variously and 

 marvellously creative power has worked in these minute 

 organisms, always with the same end in view. 



The especial function of the plumules of the Pieridae 

 was suggested in my former paper to be that of air- 

 vessels, giving buoyancy to the insects ; and these Lycsena 

 scales, by their balloon shape, are eminently fitted for 

 this service, and even in a greater degree render it 

 probable that certain Lepidoptera possess at least two 

 kinds of scales, performing different of&ces in the eco- 

 nomy of the insects. These plumules are attached to 

 the wings by an apparently hollow peduncle. They show 

 strise-like ribs, suitable for binding, strengthening, and 

 distending or contracting their balloon-like forms ; these 

 ribs are more or less beaded or articulated, by which dif- 

 ferent scales are bound or bent in various ways. The end 

 opposite to that of insertion is closed or covered with ap- 

 parently ciliary apparatus ; and they lie in rows between 

 and under the ordinary scales, which may therefore be 

 elevated or depressed at the pleasure of the insects by the 

 regulated inflation of the plumules. They differ in separate 

 species in every conceivable way — in form, in the number 

 and articulation of the ribs, in transparency, in size, and 

 in the length and shape of the peduncle ; and among them 

 are found some very anomalous forms, as in the plumules 

 of the Pieridae. In that family, Pieris agathina possesses 

 an abnormal and unique form ; and so, in the Lycaenidse, 

 Lyccena bcetica resembles it in this peculiarity. 



And as in the Pieridse, so in the Lycaenidse, it is only on 

 the males that these scales are found; it is probable that 



SER. III. VOL. III. K 



