166 R. A. DALY THE EVOLUTION OF THE LIMESTONES 



perhaps, be referred to the development of the fishes during the early 

 Devonian. This development doubtless began in relatively shallow water, 

 and the flesh-eating and scavenging fishes must have aided greatly in 

 preventing the decay of animal matter on the bottom of the extensive 

 Devonian epicontinental seas. During the Carboniferous, and yet more 

 wholesale Permian and post-Permian emergence, the fishes were driven 

 out into deeper water, where they continued the gradual colonization of the 

 entire sea-floor. So far as the fishes are concerned, that colonization 

 may have been complete in Cretaceous time. 14 That, at any rate, it was 

 complete probably several million years ago seems evident from the chem- 

 istry of the present ocean. According to Murray, the calcium sulphate 

 now dissolved in the ocean could be introduced by existing rivers in about 

 600,000 years. Since the sulphate is being rapidly decomposed by lime- 

 secreting organisms and converted into deposited carbonate, it is probable 

 that much more than 600,000 years have elapsed since the bathybial 

 fishes and other scavengers colonized the general sea-floor to depths of 

 2,500 fathoms. The test case of the Black sea shows that the present 

 content of calcium sulphate in ocean water would be largely and rapidly 

 diminished if the scavenging system were not now at work in the ocean. 

 The ratio of calcium to magnesium in the Ottawa river, the best avail- 

 able type of rivers draining the average pre-Cambrian terrane, is 3.69 : 1. 

 The ratio for the Saint Lawrence, which is not far from representing a 

 type of the rivers which might drain the average late Paleozoic terrane, 

 is 4.44:1. The ratio for the Mississippi at Memphis, similarly a fair 

 type of river draining average Triassic, Jurassic, or Cretaceous terranes, 

 is 2.50:1. The ratio for the Mississippi at New Orleans, a chemical 

 world type for the present time, is 3.92 : 1. The ratio for forty-four 

 existing rivers is 4.18 : l. 15 It appears, therefore, highly probable that 

 the ratio of calcium to magnesium for the world's entire river system has 

 been fairly constant from the pre-Cambrian to the present. We have 

 seen that this ratio is almost identical with that in the average pre- 

 Devonian limestone, but is much lower than the ratio for the Devonian 



14 This speculation regarding the migration of the fishes into bathybial and abyssal 

 depths is little better than a guess, but it is stated partly to render the hypothesis 

 somewhat more concrete and therefore more intelligible. Meager as are the relevant 

 facts concerning the fishes, those bearing on the Paleozoic and Mesozoic history of the 

 bathybial and abyssal crustaceans, echinoderms, worms, and other scavenging species 

 are almost nil. The profound mystery covering this subject does not, however, affect 

 the general hypothesis favoring a nearly limeless ocean in pre-Cambrian time ; for it is 

 next to certain that the more efficient scavengers of the sea-floor, being all relatively 

 high types, were not abundantly developed in Cambrian and pre-Cambrian time. 



15 So far as this ratio is concerned, a single analysis of a river may have high value 

 in the discussion, since Dubois has shown that, no matter how much the absolute 

 amounts of solute in a river may vary throughout the year, the proportions of the dif- 

 ferent salts remain nearly unchanged (E. Dubois, op. cit, p. 48). 



