416 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3D Ser. 



The gizzard of many more Odonata must be studied 

 before it will be safe to draw phylogenetic conclusions from 

 its armature, and for this reason I prefer not to offer any 

 suggestions on this subject based on the present results. I 

 may remark that Dr. Ris has said that Lestes appears to 

 him to be "farther removed from the typical Agrioninae 

 than even the Pseudostigma group [to which Mecistogaster 

 belongs] , perhaps even than the majority of the Caloptery- 

 ginas." To me it seemed that the principal character of 

 Lestes, viz., that the median and subnodal sectors part from 

 the principal sector much nearer the arculus than the 

 nodus, was a strong reason for regarding this Legion as 

 nearest to the Calopteryginae. Nevertheless, one of the 

 soundest results which Dr. Ris has obtained from his 

 studies is, I think, the evidence for the close relationship 

 between the Cordulegasterinas and the Libellulidas, and of 

 this evidence the gizzard furnishes an important part. 



As to the function of the gizzard in those species in 

 which it is provided with large teeth, I have attempted to 

 form an idea by comparing the size of the fragments in the 

 alimentary canal posterior to the gizzard with fragments 

 taken from in front of this organ. Such an examination 

 made under the microscope in the case of Argia agrioides 

 has failed to show any appreciable difference in the size of 

 the fragments composing the two samples. It may be said, 

 however, that the particles taken from behind the gizzard 

 were less compactly massed than those from in front, and 

 the teeth of the gizzard may have had a part in producing 

 tMs result. Further, if the teeth of the gizzard really do 

 aid in comminuting the food, one would not be surprised to 

 find that the mandibles and maxillae of species with toothed 

 gizzards were less tuberculate and ridged than in species 

 where the gizzard is not toothed. A comparison of the 

 mouth-parts of Argia agrioides and Anax Junius, as repre- 

 sentatives of these two conditions of gizzard, respectively, 

 does not seem to bear out this expectation, as the only 

 greater complexity possessed by the latter appears to be 

 two long spines on the maxilla?, wanting in the former. 



Philadelphia, 



Dec. 17, 1897. 



