124 Alexander Petriinkevitch, 



drawn out to a sharp point, almost ha\nng the appearance of a spine. 

 Anterior five or six tergites procurved, with heavily thickened pos- 

 terior edge. Pedipalpi short, pediform. Sternum long, divided into 

 three areas of which the middle one is hexagonal, the other two 

 pentagonal. Ventral surface of abdomen divided into three fields by 

 two longitudinal lines. Genotype A. rotundatus Scudder. 



Key to North American Species of Architarbus. 



1. Cephalothorax slightly produced posteriorly. Size very large 



A. horribilis 

 + cephalothorax produced posteriorly into a long lobe. Size 

 much smaller 2 



2. Anterior 6 tergites very short in the middle. Abdomen broadly 

 rounded 



A. minor n. sp. 

 + 4th, 5th and 6th tergites not conspicuously short in middle, 

 with almost parallel edges. Abdomen oval, widest in front 



A. rotundatus. 

 The specimens of Architarbus, like all other Phalangiotarbi, have 

 certain pecuHarities of structure which one must bear in mind to 

 avoid false interpretation. Both surfaces are usually superimposed, 

 the ventral over the dorsal on the obverse and the dorsal over the 

 ventral on the reverse. What is a ridge on the obverse appears as 

 a groove on the reverse. Tliis and the fact that the demarcation 

 lines between the tergites and sternites do not coincide, help to recognize 

 the structures belonging to the dorsal surface from those belonging 

 to the ventral. Even Pocock seems not to have escaped an error in 

 drawing the tergites divided by tw^o longitudinal lines into three fields 

 in some of his species. These lines belong undoubtedly to the ventral 

 surface. Thus in some specimens of A. rotundatus, where a super- 

 imposition of the surfaces for some reason did not take place, the lines 

 appear only on the ventral surface. The anal operculum is always 

 better visible on the dorsal surface than on the ventral where it be- 

 longs. The reason for tliis may be sought in an internal ring-shaped 

 thickening of the opercular edge, or still more probably in an internal 

 thickening of the wall of the anus itself. The anus itself being placed 

 ventrally, the internal ring-shaped thickening would naturally be 

 better visible on the dorsal surface after the two surfaces came in 

 contact with each other under the pressure of the drying mud in 

 which the specimen was imbedded. 



