HABITS OF THE ONUFTlIDJE. 307 



accidentally tlie water was discharged before the syringe 

 could be got into position. Upon trying to refill the 

 syringe by slowly drawing water through the Hyalinoecia 

 tube, the effort proved ineffective, excepting when the 

 worm, which was occupying the central portion, 

 occasionally moved forward. A few drops of water then 

 entered the syringe, whereupon the animal quickly 

 retreated, and the supply instantly ceased. Patiently 

 waiting and watching, this experience was repeated many 

 times, and drop by drop the syringe was filled, when the 

 worm was ejected, the protective valve at the syringe end 

 having been broken by it. This intermittent action, of 

 course, suggested the existence of hidden valves in the 

 tube, and these, when sought, were easily found. They 

 open readily in response to a forward movement of the 

 animal, or of the water inside the tube, and are closed 

 instantly by the inrush of the surrounding water on the 

 worm's retreat. This action of the valves is illustrated 

 diagrammatically by fig. 2, representing the anterior, and 

 figs. 3 and 4, the posterior end of the tube. In figs. 2 

 and 3 the arrows indicate the inrush of' water closing the 

 valves ; pressure in the opposite direction, of course, opens 

 them. Fig. 4 shows the position of matters when, having 

 reversed, the worm emerges at the smaller end, as 

 indicated by the dotted line and arrow. It will be noticed 

 that the valves (V, v", vJ") are opened, the membrane bulg- 

 ing in the opposite direction. 



The tube of l-Iyaliuoecia is evidently formed by a secre- 

 tion of the worm, and lengthening takes place by addition 

 at the wider end. This extremity is always much thinner, 

 and paler in colour. As the animal grows the tube is 

 lengthened and widened proportionately, and it seems 

 most probable that, in opposition to Johnson's suggestion, 

 the worm dwells in one tube ail its life, the portion 



