80 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



Another fact considered was tliat the hirva of Mcdanchroia {M. eephise and ^^. geometrokles), 

 formerly associated with tlic Litliosiidie, lias been shown by Dewitz to be geonietrids. Auotlier is 

 the absence of a pair of legs in the Nolidas which I find must, by their pni)al and other characters, 

 be regarded as a distinct family from the Lithosiidse. Still another fact is the couclusion I have 

 arrived at that the Lithosiida; have almost directly descended from the Tineidre or from an extinct 

 group closely allied to t1iem, and that from the Lithosiidie have arisen not only the DioptidiB, 

 perhaps including Phryganidia, the Cyllopodids, and Hypsidie, but also the Syntonjidffi and 

 Nyctemerida% as well as the ArctiidiB. 



On reexamining the larva, pupa, and imago of Phryganidia (we have no knowledge of the 

 transformations of the genuine Dioptida> as at present limited), it has seemed to me that the 

 genus has little of fundamental value to separate it from the geometrid moths. 



First, as to the larva of Phryganidia, while in the shape of the head and the slender cylindrical 

 body it (lifters little from the larva of ^Melanchroia and that of geometrids in general, if the two 

 anterior pairs of abdominal legs were atrophied there would be no essential difterence. That this 

 is probable is seen in the larva of Nola, which has but four pairs of abdominal legs, one pair being 

 atrophied. 



The end of the body (eighth abdominal segment) is humped, but the larvit; of the East Indian 

 Eusemia and Hypsa are also humped at the end of the body. Phryganidia only differs in being 

 slenderer and without hairs, and seems more closely allied to the larvae of the Hyi^sidae than to 

 that of any of the allied groups. It does not spin a cocoon. 



The puiia is obtected, and in its essential features more like those of geometrids than those of 

 Lithosiidre or any Zygipnid or Syntomid genera. It is naked and suspended by a remarkably long 

 cremaster; the end of the abdomen is otherwise peculiar. The liead i)resents no vestigial 

 characters, there being no traces of maxillary palpi, of paraclypeal pieces, or apparently of hibial 

 palpi (fig. 46). With a comi)lete knowledge of all its stages, it is still difflcult to assign it a definite 

 position. AVhen we know more about the Dioptid;^, where it probably belongs, the problem may 

 approach a solution, but that its affinities are closely with the Geometrida' is shown by comparing 

 the ])upa with that of Cleora. In the general shape of the head, of the eyes, of the front, and 

 especially of the abdomen, the resemblance is close; the peculiar shape and markings of the last 

 three abdominal segments ai'e nearly identical in both genera, though the cremaster of Cleora is 

 much shorter. 



In this connection reference should be made to the striking resemblance between the pnpte 

 of (Eta aurea and Cleora pulclirarla. To my great astonishment I find the pupa of Cleora has 

 the same vestigial head-characters as Q3ta; the general shape of the pupa is the same; the mode 

 of dehiscence the same, the shape of the vertex and its mode of separating when the moth issues 

 from the pupa case; also the same shapeof the eyes, of the peculiar cly])eus and labrum, while the 

 more pronounced vestigial characters are the labial palpi, forming a triangular area, and the large 

 semidetached paraclypeal pieces. Cleora shows that it is a more modern form in having no 

 traces of a vestigial eye-collar (maxillary palpi) such as occur (though very slightly developed) in 

 CEta. The shape of the end of the body, with the cremaster, is much the same, the shorter 

 cremaster of Cleora being an adaptation to its life in a slight openwork cocoon. In the peculiar 

 markings of the eighth and ninth abdominal segments Cleora is more like Phryganidia. 



Judging by the pu]ial characters, then, the Geometridte have directly descended from the 

 Lithosiidie, the latter, as I have satisfied myself, having directly originated from the generalized 

 Tineina. 



The imago of Phryganidia appears not to dift'er much from those of the Dioptid;^, to which it 

 has been referred by Butler. I am unable to see any important differences between the Dioptid;e 

 and CyllopodidiP, though my material is scanty. In the slender body, shape of the head, and 

 proportions of the clypeus, shape of aiitennre and palpi, both of these families do not essentially 

 differ from Melanchroia, which is now known to be a geometrid, nor from the geometrids 

 themselves. 



In its venation Phryganidia is nearly identical with that of a Josia from Jalapa, Mexico, in 

 my collection; the peculiarity is the origin of veins II2 and IIl.i from a common stem, in which 

 Phryganidia apparently differs from some if not all other Dioptid*. But the venation of the 



