OF WASHINGTON. 223 



species recently collected in the undisturbed humus of dense trop 

 ical forest at Mt. Coffee, Liberia. The habit is that of Japyx, 

 the most conspicuous difference lying in the possession of a pair 

 of jointed stylets instead of the caudal forceps. The abdominal 

 legs are the same in number and location as those of Japyx, but 

 differ in being of delicate structure and in bearing several hairs, 

 one of which has a distinctly thicker base and is slightly curved ; 

 it is attached in a manner which warrants the suspicion that it may 

 represent the claw. The antennas have about twenty (18-23) 

 joints, which are shorter and proportionately broader than those 

 of Japyx. The legs, mouth-parts, and cephalic sclerites are those 

 of Japyx throughout, but the head is not so square in front as in 

 Japyx, being narrowed so as to be somewhat triangular in gen 

 eral shape. The last segment is short and broad, the long and 

 strongly chitinized terminal sclerite of Japyx being evidently a 

 specialization necessary to make the forceps effective. 



In color the living insects were pure white or slightly creamy, 

 sometimes slightly discolored by the dark contents of the alimen 

 tary canal. There was considerable difference in size and length 

 of stylets, probably due to age. My largest specimen is about 

 3.5 millimetres long, not including antennas or stylets, the latter 

 being about .7 of a millimetre long, the former one millimetre. 

 The habits and movements are so exactly those of Japyx that the 

 first specimen was pursued as an early stage of that form. Several 

 specimens were collected, though it was far less common than 

 Japyx or Campodea in the same habitat. 



The jointed stylets might betaken as indicating relationship 

 with Campodea rather than with Japyx, but the approximation to 

 the latter in the characters mentioned above is against such an 

 inference, and the stylets, when examined more closely,* appear 

 quite as different, to say the least, from those of Campodea as 

 from those of Japyx. In the former type they are very long and 

 slender, being composed of numerous joints several times longer 

 than broad, each with several whorls of barbed hairs. The ar 

 ticulation of these rod-like stylets is rather imperfect. They are 

 only slightly flexible and very fragile. The stylets of Projapyx, 

 on the contrary, are composed of 10 distinct joints, of which all 

 except the last bear two whorls of long, simple hairs. The last 

 joint has as a distal projection a narrowly conic, truncated, trans 

 parent and rather faintly chitinized structure which may be a 

 modified claw or sense-organ, nothing comparable appearing on 

 the stylets of Campodea. 



This new form may perhaps claim the distinction of being the 

 most primitive insect yet known, since both Japyx and Campodea 

 are somewhat more specialized. Whether it should be looked 



Plate II, compare figs. 5a and 5b with 6a. 



