STRUCTURE OF SHELL. 37 



the construction and enlargement of the whole shell in 

 Balanus, should long have been viewed as a difficulty. But 

 the radii, from growing against rectangular indentations, 

 or rather furrows, in the opposed compartments, come to be 

 set a little inwards ; and their external surfaces assume a 

 very different appearance from the wall-portions of the 

 compartments, which grow against the surface of attach- 

 ment. In different species, the summits of the radii (and 

 of the alse) grow either very much more obliquely than 

 in the species figured, or more squarely — that is, they 

 extend from tip to tip of the adjoining compartments, 

 parallel to the basis. In this latter case, and when the 

 surfaces of the radii differ considerably in appearance from 

 the walls, as in Balanus tintinnabulum (Plate 1), I am not 

 at all surprised that the radii should have been described 

 as separate elements, and called " arese interjectse," or 

 " compartments of the second order :" for the shell of this 

 Balanus seems to be composed of six wedges with their 

 points upwards, namely, the parietal portions of the com- 

 partments, and of six other narrower wedges, the radii, with 

 their points downwards; and the fact that these latter 

 wedges consist simply of the sides of the parietal portions, 

 modified by growing against the adjoining compartments, 

 is completely masked. I should add, that sometimes the 

 radii are not developed, which simply means that the over- 

 lapping lateral edges of the compartments have not been 

 added to during growth. 



The alas are originally developed at the period of the meta- 

 morphosis, as slight lateral protuberances in the upper part 

 of the compartments ; from being overlapped, and therefore 

 not exposed to external influences* and from growing (as in 

 the case of the radii) against rectangular indentations or 

 furrows in the adjoining compartments, they generally 

 assume an extremely different appearance from the parietes, 

 and might naturally be thought to have a very different 

 nature. But the alae in all cases (as is obvious in PacJtt/lasma) 

 are nothing but the protuberant lateral edges of the com- 

 partments, rendered thin and modified during growth. In 

 order that the margins of the alae should be received in an 

 indentation, the upper internal surfaces of the walls of the 



