150 The American Naturalist. [February, 



thing which is not there. What he states io his note that bears on my 

 definition is, that the quadrate in Chlamydosaurus does not articulate 

 with the exoccipital, and he gives a figure to substantiate his opinion. 

 I would have preferred to have seen a figure of this structure with the 

 quadrate in place, but I have been able to confirm the observation by 

 the examination of the skull of another Agamid, Phrynocephalus oli- 

 vierii. Here the quadrate articulates with the paroccipital (supra- 

 temporal., Boul., squamosal, Baur), and is in contact with the supra- 

 temporal (squamosal, Boul.), and little or not at all with the exoccipi- 

 tal. How far the family of the Agamidse generally present this struc- 

 ture I am unable to say, as but few skeletons of this family are at my dis- 

 posal. As regards other families I have examined abundant material, 

 as stated in my last communication (1. c, Nov. p. 1004). In review I 

 may say that it is self evident that in general the distinction that I 

 have drawn betweeen Lacertilia and Ophidia in this respect is valid, 

 (1st) because the supratemporal is frequently absent, and therefore not 

 diagnostic ; (2d) because the extremity of the paroccipital is insignifi- 

 cant, and affords insufficient support ; (3d) because the paroccipital 

 process of the exoccipital is the only remaining element sufficient for 

 the purpose. 



Dr. Boulenger next attacks my definition of the Ophidia, alleging 

 that the quadrate articulates with the petrosal (prootic) or with that 

 element and the exoccipital. Here again I am supposed to have stated 

 that the quadrate does not articulate with the (prootic) petrosal, when 

 in fact, I did not mention that element. As to articulation with the ex- 

 occipital, I do not consider that this can be regarded as established until 

 the embryology and paleontology are looked into. Because an element 

 cannot be seen in an adult skull, it does not follow that it does not ex- 

 ist. The paroccipital is present in the Tortricidse, and it is, so far, 

 only an assumption to suppose that it is not represented in the allied 

 Uropeltidse, and in the less allied Epanodonta and Catodonta. 



The decurvature of the parietals and frontals to the basicranial axis 

 in snakes has been cited since Muller, by Huxley, 9 as peculiar to that 

 order, and I know of no exceptions so far as regards the parietals. 

 The optic foramina in some snakes with large eyes are confluent, as I 

 have long been aware, and this foramen is at the expense of the infe- 

 rior part of the frontals. This, however, does not produce the charac- 

 ter of the Lacertilia, and the definition is not invalidated, as Dr. Bou- 

 lenger alleges ; nor is it by the decurvature of this bone and the pari- 

 etal iu the Amphisbsenia, where they do not reach the sphenoid. I 

 9 Anatomy of Vertebrated Animals, p. 203. 



