HEKPETOLOGY OF JAPAN. 49 



As a general proposition it may be stated that there is but little 

 divergence among authors with regard to the names of classes and 

 orders. Common usage has in most cases been too strong for the 

 extremists, who have either tried to apply names according to priority 

 or who have inflicted new ones every time the '^-ank" of the group 

 was changed or a trifling component part shifted from one group to 

 another, or whose fine linguistic sensitiveness compelled them to 

 improve the name and render its form ''classic." There are names 

 which it has been impossible to dislodge, which are employed almost 

 unanimously. These it would be folly to attempt to disturb. 



The principle which has guided me in selecting the names of orders 

 and classes may be stated briefly as follows: In cases where a term 

 has not received a practically universal acceptance the oldest known 

 group name (above a generic name and later than 1758) is applied, 

 wliich covers approximately the same aggregate of genera, irrespec- 

 tive of the comparative rank assigned to the group by the original 

 namer, provided its form is not the plural of a current genus in the 

 same class, nor that now universally adopted for superfamilies, fami- 

 lies, and subfamilies. 



The application of this principle may be shown by a few examples. 

 Thus Salientia is adopted as being the oldest known group name for 

 the present order, because when originally proposed it covered the 

 same aggregate of genera as now, and none of the later names of the 

 same compass have received universal acceptance. On the other 

 hand, Gradientia has been rejected, although it is older than Caudata, 

 and the latter adopted, because the term Gradientia, as originally 

 proposed, embraced the saurian reptiles as well. Oppel's Squamata 

 has been accepted for the reptilian order containing the lizards and 

 snakes, although it included the crocodiles, a discrepancy too slight to 

 affect the general applicability of the term. Again, Testudinata, 

 although proposed four years later than Link's Cataphractae, has been 

 accepted in preference to the latter (after the elimination of the 

 older terms Testudines and Chelonia as conflicting with the earlier 

 or synchronous generic names Testudo and Chelonia) because of the 

 comparative currency of that term, as against the almost total 

 obscurity of the one coined by Link. Finally, the suborder contam- 

 ing the soft-shelled turtles is called Chilotse in spite of the fact that the 

 term Trionychoidea is both earlier and commonly adopted, the reason 

 for the rejection of the latter name being that its form is identical 

 \vith that of the superfamily. 



The Salientia may be divided into thi-ee suborders, viz, Aglossa, 

 Linguata, and Costata, of which the ftrst one is distinguished by having 

 the eustachian tubes united into a single pharyngeal opening and 

 no tongue, while the last one is amply characterized by the presence 



26485— No 58—07 i 



