TUB1CINELLA TRACHEALIS. 437 



first probably not much above the j 5 th of an inch in diameter, to a 

 cylinder nearly one inch in diameter: and this is not effected by the 

 growth of the radii, for the radii never reach the basis, and the basis of 

 course has to increase in diameter like the rest of the cylindrical shell. 

 The radii serve only to keep the summit of the shell wider than the 

 basis, which is the natural shape of this species ; and in large-sized 

 specimens, this purpose is sometimes aided by the parietes during their 

 downward growth decreasing slightly in width. In ordinary nearly full- 

 sized specimens, the parietes are of the same width at the top and 

 bottom, but in some large-sized specimens, as just stated, they even 

 become narrower towards the bottom ; as they grow only at the bottom, 

 one does not at first see how thev can ever increase in width, or how the 

 older shells can have acquired their present diameter. But an examina- 

 tion of young specimens, from *1 to *3 of an inch in diameter, at once 

 serves to show how the shell attains its full size and shape: for here 

 the parietes are all found to increase downwards sensibly in width, 

 though at a much slower ratio than in other sessile cirripedes ; in 

 larger, but not full-grown specimens, a similar increase can by care 

 be detected : hence by long-continued growth at the base of the shell, 

 with the removal of the upper part, a youug Tubicinella of small 

 diameter will be converted into an old one of large diameter, retaining 

 during all the time its sub-cylindrical form, with its summit rather 

 broader than its base. With respect to the removal of the upper 

 part of the shell, this seems almost constantly going on, for the summit 

 of every specimen invariably had a freshly broken aspect. The peculiar 

 structure of the sheath, which is the strongest part ot the shell, namely, 

 its division into oblique layers, separable by a slight force, doubtless is 

 subservient to the repeated breakage of the summit. In some species 

 of Tetraclita and of Balanus, gradual disintegration, of the upper part of 

 the shell is a necessary element in the growth of the animal, in order 

 that the orifice may increase in size, and here we have mechanical 

 breakage equally necessary. 



Some curious results follow from the peculiar growth of Tubicinella 

 just described. At Plate 17, fig. 3 b, we have a careful drawing of a 

 lateral compartment, together with its radius, (which latter does not here 

 concern us), taken from a shell *2 of an inch in diameter. The two 

 protracted dotted lines show the form which this compartment would 

 have assumed, if it had continued growing downwards at the same rate 

 of increase in width as hitherto. But the increase in width always 

 seems to become less and less as the shell grows older ; hence the 

 dotted lines, representing the wall after long-continued growth, ought 

 to have been drawn diverging or widening still more slowly than they 

 do. The lateral compartment in fig. 3 a is the exact size of the com- 

 partment of a large specimen nearly one inch in diameter ; in this 

 specimen the parietes, far from increasing in width downwards, had 

 commenced, as is represented, decreasing. Compartments of all inter- 

 mediate sizes between those figured at 3 a and 3 b can easily be shown 

 in different specimens. From these facts we may safely infer, that if 

 the whole growth of the compartment 3 a had been preserved, instead 

 of its upper end having been continually chipped away, it would have 



