GEOLOGY AND PALEONTOLOGY OF THE CANAL ZONE. 249 



when the relative frequency of specimens is known, permits an 

 estimate of the rate of their constructional work. 



This subject as a parrt of the problem of the formation of coral reefs 

 possesses an importance that can scarcely be overestimated, for it 

 comprises critical tests of both the Darwinian and the glacial-control 

 hypotheses. The topics in the foregoing list will be discussed seriatim. 



(a) Estimate of the relative quantity of coral to other constituents 

 in emerged formations containing reefs, if they have not been exten- 

 sively recrystallized, is relatively simple, although great precision in 

 quantitative expression is not to be expected. Tliis topic will be 

 further considered in discussing the Caribbean, Floridian, and Baha- 

 man fossil reefs. 



(b) Percentage estimates of material according to source are 

 difficult, but the results are of great value. The technique of making 

 such estimates is described in a memoir recently published by the 

 Carnegie Institution of Wasliington.^ 



(c) Here it should be emphasized that one of the postulates of the 

 Darwinian hypothesis is that the prism of material included between 

 three surfaces, namely, (1) the sea-bottom landward of the barrier, 

 (2) a surface assumed as an extension of the land slope under sea until 

 it intersects (3) a surface projected downward from the landward 

 face of the reef, is due to the presence of the reef (see figure 4, page 

 242). Proof that a barrier has formed during or after submergence 

 does not carry with it proof that the^ prism of material above indi- 

 cated is due to the presence of the reef. 



There are at least three criteria that can be applied in deciding 

 whether or no the flat between the reef and the shore exists inde- 

 pendently of the reef. They are as follows: (1) If the flat is de- 

 pendent on the presence of the reef, where there are breaks in the 

 barrier tongues of deep water should extend landward across the 

 shallow bottom of the flat behind the reef; and where there is no reef 

 there should be either a normal profile of equihbrium or an approach 

 to such a profile, showing a deeper flat than that behind the reef, be- 

 cause of the absence of an ofl'-shore wall behind which sediment 

 could accumulate ; but if the flat is independent of the reef, in general 

 it should be continuous irrespective of the presence of the reef and 

 should in places extend beyond the reef limits. (2) If the formation 

 of the flat is dependent on the presence of the reef, the reef should 

 stand on the seaward edge of the flat, that is, the flat should not project 

 seaward beyond the reef. (3) It is often possible to discover the 

 nature of the rock forming the sea floor between a barrier and the 



1 Vaughan, T. W., in collaboration with Cushman, J. A., Goldman, M. I., Howe, M. A., and others, 

 Some shoal-water bottom samples from MuiTay Island, Australia, and comparisons of them with samples 

 from Florida and the Bahamas, Carnegie Inst. Washington Pub. 213, pp. 23-J-297, pis. 94-98, 1918. See 

 espeeiallj- the article by M. I. Goldman, Composition of two Murray Island samples according to source of 

 material , pp. 249-262. 



