( 1G8 ) 



the comb strongly devcloiicii, the i-cduction df tlie coiub and its disappearance taking 

 place later in the series ol' genera. It, i.s worthy to emijliasi^se again tliat there 

 is also no comb in the HphiiiguUcae. The pul villus becomes small in a few s])ecies 

 and obliterates entirely in others. The paronychium preserves the two lobes of 

 each side in most Amhiilic'uiae ; in some the ventral lobe disappears, in a few 

 genera both the ventral and lateral lobes. While in i\n\ .Sphhigicae th^ pulvillns 

 disappears before the iiai'onychinm, we find among the Ambulicinue forms in which 

 the same is the case, and also such in which the pulvillns is preserved and the 

 I)aronychial lobes are lost, as is the case in Sphitigulus of the Acherontiine tribe 

 SjjIiiiujiiUcae. The ventral lobe of the j)aronychinm is often broader than the 

 lateral one, but it is never preserved when the lateral lobe disappears. 



The shape of the wing is characteristic of a great many Ambulicinae : in fact, 

 the insects are generally identified by Lepidopterists as belonging to this subfamily 

 more by tlie aspect, the shape and colour of the wings and body, than by special 

 strnctunil characters, and it is, therefore, (juite natural that some species of other 

 subfamilies have erroneously been brought, by ])revious authors, into the present 

 subfamily, and some members of the Aw.hulicinae placed somewhere else. The 

 irregular distal margin of the wings which gives so many forms a certain resem- 

 blance to leaves is not confined to the AmbuUciiute ; we meet again with a similar 

 shape among the Sesiinae and rJiilcim/iclinae, but never among the Aeherontiinae, 

 which is very curious. The falcate forewing of many Ambulicinae is found also 

 among the Philampelinae, Scsiiiiai', and the Choerorampinae, not among the Ar/w/v/i- 

 tiinae. These similarities have been mistaken for indications of closer relationship. 

 Msiuy Ambulicinae are protectively coloured ; the forewing assuming grey and clayish 

 tints, and some resemble more closely an even-edged lanceolate leaf by tlie develop- 

 ment of a dark longitudinal line or sliade representing tiie midrib {Pki/lloxiphia, 

 Clcoiis). Thongli bright colours may be said to prevail among the Ambulicinae, 

 there are a great number of very soberly grey-coloured species. The discal streaks 

 11- — M- ])resent on the forewing of so many Aeherontiinae do not ajipear here. 



The spines at the edges of the abdominal segments are always weak and elongate ; 

 they disappear, or rather become modified into scales, in several species, while on 

 the other hand, the underscales of the abdominal tergites are often spiniform, there 

 being in some sjiecies no large dentate scales present on the tergites. The sternites 

 have only in the lowest forms {Protam,biili/x and allies) some weak spines at the 

 edges ; the bulk of the Ambulicinae is without them. The abdomen is often curved 

 upwards as in some Sphingulicae. 



The })eculiar organ of friction on the clasper of the S and the inner edge of 

 the eighth tergite resembles closely that found among the Aeherontiinae in so far 

 as the mollified scales of the clasper are small and form a patch, differing widely 

 from the enlarged friction-scales of the Spkinyidae semanophorae. Sometimes 

 the friction-scales are reduced to dispersed hairs, or there are in their place only 

 dispersed granules. The patch is present in the lower forms {P rotam.bulgx and 

 allies) and disappears later on, all the more highly modified genera being without 

 it. The same applies to the patch of the Aeherontiinae. The regular ribbon of 

 enlarged scales at the edge of the eighth tergite, on the inner surface, most 

 conspicuous in Acanthosphinx (PI. LIX. f. 1), is not met with in other subfamilies ; 

 it is also less regular than in Acanthosphinx in most other Ambulicinae with a 

 friction-patch on the clasper, and is sometimes rejilaced by rough, long, narrow 

 scales, or is absent. 



