OCULAR AND GENITAL PLATES. 141 



are relatively numerous, three specimens, or 9 %. Of these, one has oculars I, IV and two have 

 I, V, II insert. A large series from that region would be well worth studying. 



The study of Sirongylocentrotus drohachiensis from different localities was made on this 

 species as an available form, to gather what might be the range in differential development 

 under various conditions and in different areas. The result shows that there is a very con- 

 siderable difference with locality, and the same fact is borne out by observations on other species, 

 as shown in Cidaris affinis, Arbacia pundulaia, Tripneustes esculentus, and Echinometra lucunter. 

 It opens up a line of inquiry which would be worth following further in these or other Echini. 



To ascertain what is the cause of variation is difficult, so many factors enter into such a 

 problem. The localities, however, represent certain groups of conditions, and percentages of 

 arrested or progressive variants have a certain approximation to those groups, so that tenta- 

 tively they may be ascribed to these conditions. The most primitive series occurred in deep 

 reentrant fjords, or bays, well removed from the open sea, which receive a considerable incre- 

 ment of fresh water. Such are Gullmar Fjord and Pulpit Harbor. The next more advanced 

 group occurred near the mouths of large rivers, St. Pierre, Dumpling Islands, and Calderwood 

 Island. Up to this point the arrested variants are in all localities more numerous than the 

 progressive. The third group occurred in open sea water well removed from large rivers, the 

 Truro locality being the purest open ocean conditions of the group; such are York, Puget 

 Sound, Massachusetts Bay, Frenchman's Bay, South Harpswell, and Truro. In all of these 

 the progressive variants are more numerous than the arrested variants. The last group is 

 far northern, and as far as known, from pure open ocean conditions, Labrador, Iceland and 

 Faroe Islands. Here the progressive variants are markedly in advance of arrested variants 

 and also of the localities farther south. 



In the series considered a curious relation is brought out in regard to which plates meet 

 the periproct when three are insert. The normal progressive variant for the species, family, 

 and order is for oculars I, V, IV to be insert, and the equivalent right-handed aberrant is I, 

 V, II. If the I, V, IV is rare, then the I, V, II is relatively frequent, as in the Dumpling Islands 

 series of 5 to 20 mm. in diameter; or in the adults of that locality, Gullmar Fjord, and Pulpit 

 Harbor. On the other hand, if the I, V, IV is relatively more common, then the I, V, II insert 

 becomes much less frequent, as seen in the series from Frenchman's Bay, Truro, and Labrador. 

 This same relative preponderance is brought out also by other species. A striking case is 

 Sphaerechinus granulans, in which (p. 126) I, V, IV is rare, 1 %, whereas I, V, II is common, 

 8 %. On the other hand, in Toxopneustes atlanticus, I, V, IV is common, 28 %, but I, V, II 

 is rare,- only 1 % (p. 122). 



Returning to the consideration of other species of Echini, in Sirongylocentrotus euryihro- 

 granimus, a southern species from Australia, of 56 specimens, 93 % have oculars I, V insert. 

 As an arrested variant, 2 % have ocular I only insert and, as progressive variants, 5 % have 



