138 THE SHAPES AND SIZES OF ANIMALS 



actual embryological development of a starfish or 

 other Echinoderm, we find difficulties in the way of 

 the view sketched out above, difficulties which have 

 not yet been overcome and which may not improb- 

 ably prove fatal. 



A starfish lays small eggs, from which a radially 

 symmetrical larva, a gastrula, is developed. This 

 larva however soon acquires a very marked and 

 unmistakable bilateral symmetry ; ventral and 

 dorsal surfaces, anterior and posterior ends, right 

 and left sides may readily be distinguished in it, 

 and the internal organisation shows the bilateral 

 symmetry as clearly as does the external shape. 

 This bilateral symmetry is not confined to starfish, 

 but is present in the larval stages of all other 

 Echinoderms as well. The importance of the point 

 is at once apparent : it shows us that the radiate 

 symmetry of the adult Echinoderm is not directly 

 continuous with, and may indeed not be the same 

 thing as, the radiate symmetry of the early larva, 

 for between the two radiately symmetrical stages a 

 bilaterally symmetrical stage is intercalated. If 

 the developmental history of an Echinoderm is a 

 true recapitulation of the pedigree of the race, then 

 the history can only be interpreted as meaning that 

 Echinoderms are descended from bilaterally sym- 

 metrical ancestors, and that the radiate symmetry 

 of the adult Echinoderm is secondary, and of later 

 origin. The matter is one of great interest, espe- 

 cially when we bear in mind that the relations of 

 Echinoderms with other groups of animals are at 

 present entirely unknown to us, and that conse- 



