THE GENERAL SHAPES OP ANIMALS. 197 



shores but on flat sea-bottoms, and some of them only on 

 bottoms of sand or mud. Here, there is none of that distri- 

 bution of surfaces on all sides which makes the spheroidal 

 form congruous with the conditions. Having to move about 

 over an approximately-horizontal plane, any deviation of 

 structure arising accidentally which leads to one side being 

 kept always foremost, will be an advantage: greater fitness 

 to function becoming possible in proportion as function 

 becomes fixed. Survival of the fittest will therefore tend to 

 establish, under such conditions, a form that keeps the same 

 part in advance a form in which, consequently, the original 

 radial symmetry diverges more and more towards bilateral 

 symmetry. 



250. Very definite and comparatively uniform, are the 

 relations between shapes and circumstances among the 

 Annulosa: including under that title the Annelida and the 

 Arthropoda. The agreements and the disagreements are 

 equally instructive. 



At one time or other of its life, if not throughout its life, 

 every annulose animal is locomotive; and its temporary or 

 permanent locomotion, being carried on with one end habitu- 

 ally foremost and one surface habitually uppermost, it fulfils 

 those conditions under which bilateral symmetry arises. 

 Accordingly, bilateral symmetry is traceable throughout the 

 whole of this sub-kingdom. Traceable, we must say, 

 because, though it is extremely conspicuous in the immense 

 majority of annulose types, it is to a considerable extent 

 obscured where obscuration is to be expected. The embryos 

 of the Tubicolw, after swimming about a while, settle down 

 and build themselves tubes, from which they protrude 

 their heads; and in them, or in some of them, the bilateral 

 symmetry is disguised by the development of head-append- 

 ages in an all-sided manner. The tentacles of Terebella are 

 distributed much in the same way as those of a polype. The 

 breathing organs in Sabella unispira, Fig. 2GO, do not corre- 



