OWENIA FUSIFORMIS. 361 



tip, the neck, which is narrowed, having a forward bulge anteriorly below the hooks, 

 and with a shoulder as it joins the shaft, which is long and slightly tapered posteriorly 

 to a delicate thread. 1 The hooks of the specimens from St. Andrews are somewhat larger 

 than those of the more bulky annelids from Naples, but precisely of the same structure, 

 and thus differ from the somewhat abrupt posterior end at its junction with the ligament, 

 as shown by Arnold Watson, the figure having probably been foreshortened to suit the 

 plate. All have a long shaft gradually tapering toward the base and a distinct shoulder 

 grasped by the epiderm. The absence of a shoulder in the figure of Sars may have been 

 due to the imperfection of his microscope. The tori of the first four segments are red from 

 their vascularity (Watson), and the anterior ones almost meet in the central line ventrally, 

 and reach as far as the bristles dorsally, so that they form the greater part of a ring. 



The tube, which varies from 3 to 10 cm., and 2 or more mm. in diameter, as a rule 

 consists of a thick internal lining of secretion, Avhich in the centre presents a circular 

 lumen in section, and of two conical, elastic membranous ends, each with a minute 

 aperture. The body of the tube is densely covered with fragments of shell, more or less 

 set on edge, and in the preparations generally sloped slightly toward the anterior end of 

 the tube. The method by which the annelid accomplishes this is graphically described 

 by Arnold Watson. The fragments are sought out on the bottom by the bilobed tips 

 of the branchias, to the mucus on which they adhere or are grasped by them and 

 " worked into the horse-shoe-shaped internal hollow, which, as a ciliated channel, conveys 

 them to the base of the crown " and subsequently to the " Lippen-organ." If suitable 

 it is first rubbed or licked by the latter, then the thin edge of the fragment is turned 

 upward, the " Lippen-organ " rises with it until its bilobed extremity projects between 

 the ventral tentacles, and the fragment is placed on the projecting pouch of the ventral 

 triangular area behind the spout-shaped process. By the action of the " Lippen-organ " 

 and this projecting triangular area the fragment is placed vertically on the elastic 

 conical sheath of the tube, glued thereto by cement from the glandular " Lippen-organ." 

 It would appear that the external thickening of the membranous tube is due to the same 

 secretion, whilst the sheath itself is formed from within, as the imbricated arrangement, 

 for instance, of the posterior end clearly shows. Watson noticed that the secretion of 

 the first two pairs of thread-glands and the long cells of the " Lippen-organ " strongly 

 resist stains, whereas the secretion of the thread-glands of the posterior region stains more 

 easily. The annelid can readily reproduce the anterior conical cap of the tube by the 

 secretion of the thread-glands of the anterior region as well as the glands of the " Lippen- 

 organ," and it also frequently cuts off portions of the tube by acting on the inner membra- 

 nous portion and by aid of the tentacles and "Lippen-organ" bursting it off. Watson also 

 points out that the tube, though fairly tough during life, speedily decomposes after the 

 death or expulsion of the animal, which usually dies when removed from it ; yet those in 

 the stomachs of fishes remain as tough as in life. A ball of sand, loosely held together 

 by mucus (probably secreted by the tentacles), generally invests the naked tube 

 posteriorly. The position of the shell-fragments on the tube offers great resistance to 

 the raising of the tube in the sand, hence the necessity for considerable muscular exertion 



1 De St. Joseph calculates that there are about 7,600 hooks per torus, and perhaps about 

 450,000 in all. 



169 



