LEFEBVRE ON MANTIS. 71 



the Orthoptera always to devour one another, as has some- 

 times been accidentally observed among the Mantides. This 

 question remains, therefore, to me insoluble, and this express 

 condition of living in the most uninhabited and most uninhabit- 

 able places is, to me at least, incapable of explanation. But if 

 the habitat of these insects attracted my attention in some 

 particulars, the organic conformation of one of them was not 

 less able to fix it most intently. 



Up to the present time, all authors have agreed in recognising 

 five articulations in all the tarsi of the Mantides ; and yet one 

 of the individuals which I found among them exhibited four 

 only on the anterior, and three on the intermediate and 

 posterior legs ! 



Although they were in the pupa state, it is not to be sup- 

 posed that the development of the other articulations takes 

 place at the time when these insects arrive at perfection, since 

 the larvae of species allied to them, as in all the other known 

 Matdides, have five articulations in all their tarsi. I could not, 

 with the most powerful microscope, detect even the rudiments 

 of the missing joint, which might have been attached to the 

 adjacent part, as is observed in some insects. 



With respect to this anomalous and puzzling conformation, 

 I should have been tempted to consider it as one of those 

 whimsical freaks of nature which sometimes occur, had it 

 not been for certain characters peculiar to this insect, and 

 which I shall point out in referring to this species, which 

 afforded a most marked difference between it and the other 

 Eremiaphilcs ; in short, if in the work on the Expedition to 

 Egypt — (Pi. 2, fig. 5,) I did not find this very insect accurately 

 figured, and (fig. 6. d,) this same anomaly faithfully portrayed. 

 It is not probable that, after a lapse of thirty-four years, the 

 same monstrosity should have reappeared. Laying aside this 

 supposition, which cannot reasonably be admitted, it must be 

 allowed that this species has in fact but four and three articu- 

 lations of the tarsi, and that there may be other species of 

 Mantis of a similar conformation. 



Reflecting upon the recent observations on the number of 

 the articulations of the tarsi in Coleoptera, and their disputed 

 importance in classification, it must be remarked that in this 

 anomaly of the tarsal joints nature has only followed the line 

 she has pointed out in the heteromerous Coleoptera, where the 



