ON COKAL ISLANDS. 229 



horizontal mode of growth not checked in any way, we find 

 every grade of transition. The precipitous fall is more or less 

 oblique as the stream impinging on the reef is more or less 

 strong.'"* 



It is my conviction, derived from my own long-continued 

 study of coral reefs, that the connection here indicated between 

 the strength and direction of the stream on the one hand and the 

 vigour of growth in the corals and in the reefs they form on the 

 other, is one of the principal causes that have given the reefs 

 their frequently very remarkable forjns. This view is in c'ireot 

 contradiction to Darwin's theory-, which now finds universal 

 acceptance, as to the formation of coral reefs. The reception 

 which that has met with, as well as the high respect and vene- 

 ration which I am most ready to pay to every opinion uttered 

 by Darwin, compels me to devote the ensuing chapter to a de- 

 tailed description of the reefs of the Pelew Islands in the Pacific, 

 which during ten months I thoroughly investigated. Every 

 variety of reef occurs there within a limited space, and a 

 close study of their structure will show that Darwin's theory 

 is not sufficient to explain them all, but, on the contrary, that 

 in every instance hitherto investigated, whether on a large or 

 small scale, the same efiects on their growth, as produced by 

 strong and constant currents, may be detected. 



