302 PROCEEDINGS OF THE ACADEMY OF [May, 



indicate the insertion points of these hairs. A median section of it is 

 represented in fig. 25, where the apex is at the left and the dorsal 

 border uppermost. It is covered by a cuticula (Cut.), which is less 

 thick than that of the adjacent body surface, beneath which is a 

 hypodermis (Hyp.) considerably thicker than that of the rest of the 

 abdomen. The interior of the organ is a blood-space (Bl. Cav.), 

 shown by stippling, containing neither glands nor muscles, though 

 muscles insert near its base. The colulus is equally developed in both 

 sexes. Thus it is a simple integumentary extension with a large 

 axial blood cavity. An organ of its volume must surely fulfill some 

 important function. Its thickened hypodermis might suggest sensory 

 nature. But I think it may be more likely a supplementary respira- 

 tory organ, for which the combination of relatively thin cuticula and 

 vascular space would speak, and furthermore the trachese are relatively 

 poorly developed in this species. 



In Peucetia (Oxyopid) and Lathrodectus (Theridiid) it is about twice 

 as long as broad, shorter and relatively broader in Dysderids and 

 Argiopicls. But in the Agalenids I have found the most interesting 

 relations : in Tegenaria derhamii it is a plate about twice as broad as 

 long, in Hahnia bimaculata a transverse plate eight times as broad as 

 its antero-posterior length, while in Agelena ncevia there are a pair of 

 small plates completely separated in the mid-line by a little more than 

 their diameter and inserted well anterior to the roots of the anterior 

 spinnerets. The last case is interesting, for it exhibits the colulus in 

 a distinctly paired condition, the first case of such a relation yet known. 

 We have next to consider the ontogeny of colulus and spinnerets. 

 Salensky (1871) first proved that the appendages of the fourth and 

 fifth abdominal appendages become the spinnerets. He found the 

 fourth appendages originate the anterior spinnerets, the fifth the 

 posterior, and that the median spinnerets "do not develop from 

 the abdominal feet, but appear between them" (I cite from Jaworow- 

 ski, for I cannot read the original Russian). This origin of the anterior 

 and posterior spinnerets has been confirmed by Locy (1886), Morin 

 (1887), Kishinouye (1890) and Korschelt (1892), while Barrois (1878), 

 Balfour (1880) and Schimkewitsch (1887) erroneously held that the 

 embryonic appendages disappear and that spinnerets are new for- 

 mations. The most detailed and satisfactory account is that of 

 Jaworowski (1895) of Trochosa. He found that after reversion each 

 appendage of the fourth and fifth abdominal segments is bilobed, 

 consisting of an outer sacculus (exopodite) and inner sacculus (end- 

 opodite) ; on the fourth segment the exopodites become the anterior 



