144 LANICE CONCHILEGA. 



of bivalves. The coarser tubes show large fragments of shells, half a Oardium or other 

 shell conspicuously projecting here and there, and the nacreous (or it may be the 

 purple) surface of the mussel and the nacreous surface of such as Nucida, already 

 mentioned, is exposed. These fragments often slant from before backward, but with 

 frequent interruptions, so as to render the tube as difficult to draw by front as by rear 

 through the sand. All these shelly-fragments and particles of gravel would appear 

 to be placed and cemented in position after the thin inner lining of the tough secretion 

 is formed. The lining of secretion keeps the channel smooth so long as no bending 

 occurs. A tube with moderately coarse grains is still translucent here and there so 

 as to give the annelid light, if that is of importance to it. The locality in which the 

 tube is situated, of course, has an influence on its structure, those from sandy regions 

 being fine, those from gravel or shell-gravel are coarse. Those from deep water 

 generally have more of the membranous and less of the shelly materials — it may be from 

 their scarcity. The coiled tubes situated in the valves of Mytilus modiolus present 

 considerable patches of membrane, and one surface is glued to the shell, the protective 

 structures being fragments of heart-urchins, entire green-pea urchins, spines of 

 sea-urchins, and heart-urchins and small shells of Dentalium. At Lochmaddy the 

 tubes are often composed of minute fragments of gneiss and quartz, whilst at 

 Connemara fragments of Lithothamnion are common, the coarser fragments being 

 posterior, and the finer, with Foraminifera and other minute particles, forming the 

 branched termination. 



In some the anterior aperture has two flattened plates composed of scale-like 

 fragments of bivalves' or of entire valves, the fimbria? being attached to the margins, 

 and the two plates, by the elasticity of the secretion, are closely applied except on the 

 issue of the annelid. Moreover, much finer fibres are sometimes attached to one end of 

 the flattened anterior aperture. 



Occasionally the ordinary tube is again continued beyond the branching anterior 

 end, and a new arborescent termination constructed. 



The arborescent anterior end of the tube is formed by cementing such elongated 

 bodies as the spines of the heart-urchin, elongated fragments of shell or gravel, which 

 are placed parallel to the long axis of the fibre or its branches. Moreover, where the 

 tubes are composed of fine grains the fibres are correspondingly delicate. In some a large 

 fragment of shell is utilised in the arborescent tuft, apparently for strength, or in the 

 case of a limpet-shell for protection. The arborescent tuft has a transversely flattened 

 form, and the large aperture is protected by a basal web at each side (Plate CXIX, 

 figs. 8 c and Sd). Occasionally the annelid avails itself of the neighbouring branches 

 of zoophytes or Polyzoa. The arrangement of the branches at the anterior end of the 

 tube is apparently connected with the safety of the delicate tentacles and branchiae. As 

 the advancing tide covered the tubes at Lochmaddy a cloud of mud was thrown out by 

 the annelids. 



Mr. Arnold Watson, who has done such excellent work by observing tubicolar forms 

 constructing tubes in his aquaria, was good enough to send a copy of a photograph 

 of a newly-constructed branched process at the end of a tube showing graceful curves of 

 the filaments formed of single grains of sand and minute particles of shells. Subsequent 



