( 32fi ) 



which is widely separated from tlie ]iciiis-sliiatli (I'i. XXIX. f. 38. 30. 44); the 

 latter with one or two teeth. 



?. Middle segments of antenna liigher tlian long. A ridge in front of the 

 vaginal cavity strongly chitinised, mesially sinnate or cleft (PI. XV. f. 18. 19). 



Larva and pnpa essentially as in S/fhinr. — Food : Pntnufs and other trees. 



IJa6. Nearctic Region east of the Mississippi basin, in ( 'anada extending further 

 west. 



The abdomen is said by Fernald, from whom Grote and Smith have copied the 

 statement, to be without s]>inules ; in this the Professor was wrong (see above under 

 S/i/iinu^, the abdomen agreeing in spinnlation with that of Sphinx. 



We have not seen a s])ecimen of any of the three species of this genus in 

 which there was an indication of more than one blue spot in the eye-jjatch of the 

 liindwing. The only spot present varies in size and position. With which of the 

 three original spots is it homologous ? Does it correspond to the tirst, second, or 

 third spot of the species of .S)V////,y ? We tind among Spli. jamaicensis specimens 

 with only one central bine spot: is this tlie same as the spot of Calasi/mf/ul'ts'f 

 A comparison of the various sjiecies reveals the true homology. The three grey 

 lines of Sp/i. kindermnnni, sejiarated by black scaling, which re[)resent the more 

 ancestral form of the eye-patch, are ])ale bine in Sph. cerisi/i sal/ceti and remain 

 separate ; the second and third become arched, join each other in front, and form 

 a more and more complete ring in the other snbsj)ccies of ceiisiji, the first sjiot 

 disap])earing at the same time more or less. In ocfUafn, which is very closely 

 allied to reris'/i, the ring is always complete and the first spot has entirely dis- 

 appeared. The same is the case in planus. If we bear this in mind and 

 compare now the hybrid asti/ht.i x ocrllatn we observe at once that the ring of 

 this liybrid consists of the sjiot of asfi/his and the outer part of the ring of ocellata, 

 which part is the third spot of cerixip snliceti; therefore the spot of asfi/lus must 

 either be the second or the first spot of .saliceti, i.e., the spot of Ca/ns>/nil/oliis is not 

 the spot of Sph. jamaicensis f. ab. jamaicen.His (see above), the blind eye of this 

 jamaicensis and of Calasymbolus are not homologous in respect to their pupil. 

 Further, if the Calasymboliis pujtil were tlie first spot of saliccfi, in the nnion 

 astyhis x ocellata all three spots would be j)resent and should be represented at 

 least by some blue scales in the offspring of such a union. This is not the case in 

 the hybrids examined, and we conclude, therefore, that the \m\)\\ of a.-itylus (and the 

 other Calasi/mbolus) is homologous to the proximal jiart of the ring of ocellata, i.e., to 

 the second spot or line of hindermanni and cerisyi salicefi, as explained above. 



It would be very interesting to have the results of forced breeding on the lines 

 of Prof. Standfnss's experiments. Such researches would doubtless result in the 

 production of individuals of Calasymhulus with more than one blue spot in the 

 eye-])atch. 



Key to the species ; 



a. Fore- and hindwing scalloped . . . 274. C. excaecata. 

 Fore- and hindwing not scalloped . . . . b. 



b. Distal margin of forewing distinctly angnlated 



at M\ three distinct brownish lines with 



grey interspaces ^75. C. mi/ops. 



Distal margin of forewing not distiiuttly 

 angnlated at M' ; the discal lines ijnite 

 indistinct, except at costal margin . . 276. C. astylus. 



