THE SPECIES OF AEGLA — SCHMITT 449 



the lateral margins to form a very definite, rather sharp carina ex- 

 tending straight forward to the anterior extremity of the rostrum, 

 which is distally not, or at most only slightly, bent upwards. At 

 about the level of the corneae the dorsal carina of this type of ros- 

 trum almost always attains a greater height or elevation above the 

 lateral margin than the ventral keel has depth below the margin. 

 Above the level of the lateral margins the rostrum in cross section 

 is definitely triangular, like the roof of a ridge-roofed or gable-ended 

 house. The sides of this roof run straight down from the ridge 

 or carina to the lateral margin either side at about a 45° and often 

 steeper angle (that is, at about the middle of the free portion of 

 the rostrum or between that point and the level of the anterior margin 

 of the corneae). At most these lateral slopes in this first group may 

 be slightly concave ; they are, however, never particularly depressed or 

 flattened down, excavate, or longitudinally grooved or troughed. 



The other type of rostrum (Pacific or Andean) is fairly flat from 

 side to side and not as a rule at all like the so-called ridge-roofed 

 type, although some species assigned to it {A^ section of the key, 

 p. 454) may liave rather a sharp rostral carina (i. e., A. rioUinayana). 

 In general, rostra of this type in cross section form more of a flat 

 longitudinally corrugated roof than a steep-sided ridged roof, inas- 

 much as the sides of the roof either side of the median carina are 

 usually more or less depressed below the lateral margins, and ex- 

 cavate or longitudinally troughed. The height to which the rostral 

 carina rises above the lateral margins, at about the level of the 

 corneae, is usually appreciably less than the depth to which the ven- 

 tral keel extends below the lateral margins. As a general rule, the 

 dorsal carina tends to fade out or disappear as a carina before 

 attaining the distal extremity of the rostrum, which is generally more 

 or less definitely recurved or bent upward. 



A few species seem to have rostra of an intermediate or transitional 

 type that may not have been altogether satisfactorily placed in our 

 key. However, such species have been assigned to that primary 

 group, A^ or A^, to which they appeared to be most closely related, 

 all characters considered. A. jujuyana and A. franca have been 

 assigned to section A^ of the diagnostic key, and A. affinis to section 

 A^. This last-named species, in the unique specimen at hand, has the 

 dorsal rostral carina somewhat higher at the level of the corneae 

 than the ventral keel is deep, yet its basally broad and flattened 

 rostrum is certainly indicative of a nearer relation to the A- than to 

 the A^ species. A. jujuyana and A. humahuaca fall into opposed 

 primary sections of our key on the basis of the character of the 

 rostrum ; nevertheless, there is in some respects a tantalizing resem- 

 blance between the two that suggests a suspiciously close kinshij^. 



