j;;i; SAMUEL II. SCUDDER ON THE 



the Lampyridae and Heteroceridae, 1 we shall be loth to assert a close affinity with these 

 groups. Such groups of Coleoptera as have aquatic larvae show, however, no points 

 of resemblance at all to Mormolucoides, and it seems, therefore, far more probable that 

 ilicv arc neuropterous. 



In support of this view, we have on general grounds, the flattened and posteriorly ta- 

 pering form, much more common in Neuroptera than in Coleoptera, besides the ter- 

 minal cerci, and posterior lateral appendages of the abdominal segments — features much 

 more in accordance with the structure of those groups of Neuroptera to which they 

 seem most nearly related, than with the structure of any Coleoptera. 



These groups are the Perlidae, Ephemeridae and Sialidae, in all of which the larvae are 

 at least in large part aquatic In each of the first two of these groups, there is a re- 

 markable uniformity of larval organization, and thej r seem to differ so much from Mor- 

 molucoides as to make it unwarrantable for us to look for intimate relationship with 

 them. In Perlidae, for instance, we have a prothorax distinctly differentiated from the 

 other thoracic segments, and the latter bearing at a comparatively early age, as in Blat- 

 tariae, indications of the coming wings in the form of pad-like expansions of the outer 

 angle of the said margin; we have also long and prominent antennae, very long and 

 large flattened legs, anal cerci of great length, and no sign of an inner pair of cerci. 

 In Ephemeridae, we have an entirely different form, equally discordant in its relations to 

 Mormolucoides. The legs are nearly as long and stout as in Perlidae, lateral respiratory 

 filaments cover the dorsum of the abdominal segments, the head bears stout, and often 

 long antennae, while the terminal segment is almost invariably armed, not only with outer 

 large, long, feathered anal cerci, but also with a similar, single, median style, even when 

 the latter is absent from the imago; two inner styles are never present. 



The comparative uniformity of larval structure among the diverse genera of each of 

 these two groups prevents lis from believing that Mormolucoides with its very different 

 structure could by any possibility be included in either of them. Not a trace of thoracic 

 wing pads or abdominal respiratory filaments can be seen on the hundreds of specimens 

 examined. The great length and size of legs and mill Particulate antennae in both the 

 groups, find no counterpart in Mormolucoides, and the appendages of the terminal seg- 

 ment are altogether different. 



Not so, however, or not by any means to so great an extent, when we compare the 

 larvae of Sialidae. Here we find a considerable greater range of characteristics, so that 

 it is not so easy to recognize a common facies among them. But we may note one or two 

 characteristics by which they approach much more closely our fossil type. All the ap- 

 pendages, — antennae, legs and (often) the cerci, are shorter and slenderer than in the two 

 groups last mentioned. In some, the antennae at least are comparatively insignificant. 

 The mandibles in some are very stout, and though long in all that are known may well 

 be believed to be capable of modification in this regard. The abdominal segments are 

 provided with lateral filaments, projecting backward from the posterior outer angles. 

 The appendages of the terminal segment vary very much, some having a single median 

 style of considerable length, others a shorter lateral pair, in some cases furnished api- 



1 Tlic Heteroceridae live near but not in water. 



