14 



In the Cochlidids, whenever there are pectinations, and whatever 

 be their form, they are always free of scales internally and scaled 

 externally, and every portion of the antenna not occupied by pec- 

 tinations is completely scaled. This may be a scrap at the tip, half 

 the antenna, the whole of it except a little irregularity at one point 

 near the base. Or it may be, as it is in our English species, the 

 whole antenna. 



I speak with much diffidence, as my own acquaintance with 

 Cochlidid species is very small ; but so far as I understand the 

 views of Dr. Dyar and others as to the phylogeny within the family, 

 these simple antennaed species, like ours, which have also smooth 

 larva when full-grown, are the most recently evolved. If this proves 

 on further research to be correct, we have here a proved instance of 

 a pectinate antenna reverting to a simple form. 



Having done so, it retains the character of a completely scaled 

 antenna, which belonged to it throughout, but which was obscured 

 by the pectination. It retains also the irregular disposition of the 

 scales, which is a character associated with pectination. 



In Phitella {a?inu/ateila, etc.) the antenna has a complete coating 

 of hairs except for the interruption of a ring of short scales at the 

 distal margin of the segment, and of several placed dorsally repre- 

 senting the first ring. The scales are so small that the rows seem 

 widely apart, and the clothing of sense-hairs is nearly as complete 

 dorsally as ventrally, the difference being only by the interruption 

 of the two or three scales of the first row. 



In Coleophora there are two rows of scales ; the second row is 

 interrupted ventrally, giving place to hairs ; in some species this 

 interruption is so small and evanescent that the antenna appears to 

 be one with a complete coating of scales. 



In Cerostovia {xylostella) the arrangement is the same as in 

 Phitella, except that the scales are much larger and overlap, and 

 appear to have banished all hairs from the dorsal region. 



In Depres sarin, Aplota, Hyponomeiita, Acrolepia, and many other 

 forms there are two rows of scales, with the first row deficient ven- 

 trally, leaving room for hairs, much as in Coleophora. 



In CEcophora there is a ventral line of hairs from end to end of the 

 antenna, but also a few hairs exist between the scales, which cover 

 in two rows to a segment the dorsal and lateral aspects of the 

 antenna. 



In Argyresthia and Elachista the arrangement is the same, 

 except that no hairs can be detected dorsally ; so that here we 

 have, practically, the obtect antenna, in the form in which the 

 scales broadly clothe the dorsal aspect of the antenna. 



The Pyrales in all their sub-families, including Phycid-^, Cram- 

 bids, Scopariads, etc., have antenna scaled as in Obtectte. 



Orneodes has the scaling well advanced laterally, having some 

 resemblance to that of Argyresthia. It is substantially a pyrale or 

 obtect antenna. 



