32 



Tn conclusion, il seems thai (lie male genital armatnre, irrespective 

 of the intromilleut orjian, consists of a supra-anal plate, a recurved spine 

 (usually siniph> and acute), and of a pair of large, subspatulate, risually 

 simple, lateral claspers ; and that all these ']>ar(s are usually concealed by the 

 larire, scpian", and open, or conical and (dosed terminal (in the imago) segment 

 of tlu! al)domen, which is necessarily ruptured hefore the parts can he put to 

 their normal use. 



The male armature is nnudi simi)lcr in the VhalanUhx than in some 

 of liie Koclnuhv, where the lateral claspers ai'c often very complicated (as in 

 ILifkna desh-Kctor), though they are much as in Agrotis tessellata. The supra- 

 anal plate and spine arc much alike in certain genera of the two families. 

 I doul)t, however, if any reliable family (diaracters, separating the Nodukla^ 

 from the FhaJa;nUhe, can be drawn from the genital armature. 



§ 7. C'OMPAKISON WITH OTIIKU FAMILIi:S OF LEPIDOPTERA. 



Comparing the denuded head of the rinihenich^ we liave 'described 

 with a typical Noctuid, such as Mamesfra arci/ca, the eyes in the Pkalcenidce 

 are fuller, more spherical, nearer together; the clypeus much longer and 

 narrower; tiie cpicranium is a little smaller in proportion, narrow(^r, and 

 more elevated, whil(3 tlic occiput is longer in proportion to the width; the 

 labruni is rather snudler and shorter, while the maxillai arc shorter, smaller, 

 and weaker. Indeed, the shortness of the occii)ut in Mamestra carries tlic 

 whole front back, and it is less vertical than in the Phalanidce. This is seen 

 especially in I'^utrapcla and Ci/diinou, where the front of the head is much 

 longer and narrower than in the Noditidcc. 



Now, turning to the hitherto doubtt'ul genera Don/odes and Pac/ii/cncmia, 

 they agree well with the ixdations of parts in Mdiiietira. The head is much 

 alike in both of these genera ; the ocelli arc nearly as large in ])ropfirtion, 

 and as shining bhudv, as in Afdiitestra, in coku' dilfering entirely from the 

 ocelli of the riiaUnnidoi, which are almost indistinguishable in color from 

 the surliice of th(! head. In these two genera, also, the cpicranium is larger 

 than in any I'hahenid known to me, forcing liack, as it were, the occiput out 

 of sight, the latter being very short, merely a transverse rim on (he posterior 

 edge <if the head. The cly])(!us in Dori/adrs is scarcely longer than ils width 

 at the I)ase, and il is still shorter in Pachijcnciniii, with tin; edge (hickene<l 



