206 DK. W. B. BENHAM ON 



The chcetal ring is incomplete, presenting a small dorsal gap 

 «qual to about twice tbe usual inter-chaetal space ; there is no 

 ventral gap. The chaetse of segments v., vi., vii. are larger tlian 

 the rest, as in the preceding species ; wiiile those o£ segments iv. 

 and VIII. are intermediate in size between these and the normal 

 chsetse. 



There are no chsetae on the clitellum, wbicb is fully developed. 

 The chsetse number 19 on segment ii. 

 44 „ VI. 



51 „ XIII. 



52 „ XXVI. 



The number of segments in the worm varies considerably, for 

 instance : — 



One worm measuring 70 mm. bas 66 segments. 

 „ „ 110 „ 88 „ 



,, „ 120 „ 88 „ 



,, ,, 125 ,, 95 ,, 



125 „ 100 „ 

 This slight discrepancy in the relation between length and 

 number of segments suggests that the measurements are not 

 true : the worms ate so soft that as one holds them, they stretch 

 ±0 nearly any length. A worm containing 100 segments, for 

 instance, should not measure so much as 125 mm., for the seg- 

 ments are not as much as 1 mm. long, in such a relatively small 

 worm : it is more probable thaC the true length of each segment 

 is nearer | mm., and a worm with this number of segments 

 would be 75 mm. long. Such an estimation of relation of 

 number to length is derived from the study of carefully pre- 

 served specimens of other species of PerichcBta, which have not 

 undergone undue shrinkage. 



Internal Anatomy. — The study of the septa in this worm 

 suggests that the usual location of the gizzard in the genus 

 requires further investigation. The peripharyngeal septa are, 

 as usual, fairly stout ; the septum between segments vii./viii. 

 is, as usual, distinct and attached to the anterior margin of 

 the gizzard (PL 16. fig. 4 e) ; there is also a septum viii./ix., 

 .attached to the body-wall at the intersegmental groove, and 

 passing as a delicate sheet of tissue — not a mere strand, as is so 

 often the case — to be inserted in the gizzard, just before its 

 hinder end becomes everted to form the rim of the bell. 

 .Septem ix./x. is absent, but the four following septa are quite 



