64 THE entomologist's record. 



hind margin, which two rows unite together towards the outer margin ; 

 in the ? instead of this is only a short raised semi-oval curve. 

 Forewings with 11 or 12 nervures, 1 free, 2-7 at equal distances 

 from cell, 8 out of 7, 9 divided near apex of wing when 12 nervures 

 are present ; in the latter case, however, 10 is from 9, 11 from top of 

 cell, 12 free ; if nervure 9 be not forked 10 arises from front edge of 

 cell and 11 is free ; hindwings with 9 nervures, 1 free; the following 

 with tolerably equal interspaces : 4 from apex of cell, 5 as a 

 continuation of one of the weak dividing nervures of cell (the fold ?) 

 or even further on towards the costa, 6 and 7 on a short footstalk, 8 

 from base sending back an oblique nervure to middle of front margin 

 of cell." Two main divisions were then made of the Spliinf/ides: 

 I. Antenn;e with the apex pencillated, hooked. II. Antennae neither 

 pencillated nor hooked. 



The following classificatory table is then based on these two main 

 divisions : 



I. Antennffi with the apex pencillated, hooked. 



1. Abdomen with sides and anal segment barbed . . . . Macroglossa. 



2. Abdomen with sides and anal segment not barbed. 



a. Abdomen with anal segment acute . . . . . . Hphinx. 



b. Abdomen with anal segment obtuse . . . . . . Acheruntia. 



II. Antennee neither pencillated nor hooked . . . . . . Smennthus. 



This classification appears to be a very natural one, and, in fact, 

 has, with small modifications, as will be seen later, been used up to 

 the present time. The four genera have since, owing to more detailed 

 knowledge, been broken up into smaller natural genera and have thus 

 come to have subfamily value. The characters derived from the 

 neuration of the Sj)}iijii/i(le>i as used by Herrich-Schafler marked a 

 great advance. Nervure 9 of the hind-wing (the cross nervure between 

 8 and the cell) was here for the first time used as a classificatory 

 character and has proved to be most valuable and constant. 



Stainton, twelve years later, in his Manual of British Lepi/Iojitera, 

 1857, followed Stephens almost to the letter, both in his placing the 

 Zi/iiacnular, Aef/eriidae and Sesiidac with the Sjildni/idae in the 

 division Sphint/ina, and also in the generic subdivision of the 

 Sphiniiidae, except that Stainton used five as against the four genera of 

 Stephens, the addition being Vhoeroeavipa characterised by the larva 

 having the anterior segments retractile. Except for this character 

 no other is used to differentiate the genus from Deilephila. The 

 Sedidae [hoiiibi/lifonnu, fnciforiiiis, &c.) are separated into another 

 subfamily. Although nearly 30 years subsequent to Stephens, 

 Stainton failed to make any advance, indeed one must confess 

 there was retrogression, as the palpi and other structures are not 

 even used for classificatory purposes at all. Seeing that Herrich- 

 Schaffer had only so recently classified the European lepidoptera, 

 Stainton 's total neglect to utilise his system is remarkable. Eighteen 

 years after Stainton had published the Manual, Butler, in 1875, 

 brought out his Momxjraplt of the SpJiiiu/idae. In most respects this 

 was a disappointing work, c.ij., (1) The general characters of the 

 Sphinf/idae as a family are totally wanting. (2) No mention is made 

 as to the affinities of the Sphint/idae with other superfamilies, &c. 

 Six subfamilies are noted, Macroiilossiiiac, Clwerocainpi)iae, Anibiilicinae, 

 Sinerinthinae, Acherontimae, and Splmujinae. Macruylossinae \^ Qqxxdl to 



