704 



SCIENCE 



[N. S. Vol. XXVIII. No. 725 



the quantity remaining in solution in the 

 ocean, the result leaves us with the ap- 

 proximation of twenty million cubic miles 

 of matter once in solution, and now for the 

 greater part existing as precipitated or ab- 

 stracted deposits at the bottom of the ocean. 

 "We are to distribute this quantity over its 

 floor. If the rate of collection had been 

 uniform in every part of the ocean through- 

 out geological time, a depth of about one 

 seventh of a mile (240 meters) of deposit 

 would cover the ocean bed. 



While, I believe, we can place consider- 

 able reliance on this approximation, we are 

 less sure when we attempt an estimate of 

 its mean radio-activity. If we assume for 

 it an average radio-activity similar to that 

 of Globigerina ooze, we find that the quan- 

 tity of radium involved must be consider- 

 ably over a million tons. Apart from the 

 value which such estimates possess as pre- 

 senting us with a perspective view of the 

 great phenomena we are dealing with, it 

 will now be seen that it supports the find- 

 ing of the experiments on sedimentary 

 rocks, and leads us to anticipate a real dif- 

 ference in the radio-activity of the two 

 classes of material. 



The Sedimentary Bocks:— The radium 

 content of those of detrital character is in- 

 dicated in the following sandstones, slates 

 and shales: 



Shales, sandstones, grits (10) 4.4 



Slates (Cambrian, Devonian) 4.7 



Mud from Amazon 3.2 



Some of the above are from deep borings 

 in Carboniferous rocks (the Balfour and 

 Burnlip bores), ^° and from their nature, 

 where not actually of fresh-water origin, 

 can owe little to oceanic radio-activity. 

 Many of the following belong to the class 

 of precipitates, and therefore owe their 



" For these rocks, and for much other valuable 

 material, I have to thank Mr. D. Tate, of the 

 Scottish Geological Survey. 



uranium wholly or in part to oceanic 

 source : 



Marsupites chalk 4.2 



Green sandstone 4.9 



Green sand (dredged) 4.5 



Limestones and dolomites [Trenton, Carbon- 

 iferous, Zechstein, Lias, Solenhofen (7)] .. 4.1 



Keuper gypsum 6.9 



Coral rock, Funafuti bore (4)" 1.7 



Trias- Jura sediments, Simplon: 17 rocks of 



various characters 6.9 



Mesozoic sediments, St. Gothard: 19 rocks of 

 various characters 4.2 



The general mean of sixty-two rocks is 

 4.7. 



Making some allowance for uncertainties 

 in dealing with the Simplon rocks, I think 

 the experiments may be taken as pointing 

 to the result: 



Igneous rocks from 5 to 6. 



Sedimentary rocks from 4 to 5. 



If our estimate of oceanic radium be ap- 

 plied to the account of the sedimentary 

 rocks in a manner which will be understood 

 from what I have already endeavored to 

 convey, there will be found to exist a fair 

 degree of harmony between the great quan- 

 tities which we have found to be in the 

 sediments of the ocean and the impoverish- 

 ment of the sediments which the experi- 

 ments appear to indicate. 



In all these results fresh and unweath- 

 ered material has been used. The sand of 

 the Arabian desert gave me but 0.4. Simi- 

 larly low results have been found by others 

 for soils and such materials. These are not 

 to be included when we seek the radio- 

 activity of the rocks. 



As regards generally my experiments on 

 the radium-content of the rocks, I can not 

 say with confidence that there is anything 

 to indicate a definite falling off in radio- 

 activity in the more deeply seated materials 

 I have dealt with. The central St. Gothard 



" For these I have to thank the trustees of the 

 British Museum and Mr. A. S. Woodward, F.R.S. 



