360 ANNUAL KEPORT _SMITHS0N1AN INSTITUTION, 1908. 



It is remarkable, and may be significant, that in these latter rocks I 

 have reached the lowest radio-activities I have riiet — down to almost 

 one-million millionth of a gram of radium per gram; although the 

 general mean of the St. Gothard igneous rocks, owing to the high 

 radio-activity of the Finsteraar granite at the north end of the tun- 

 nel, is not exceptionally low. Eadio-active minerals seem common in 

 the Simplon rocks, involving considerable variations in successive 

 experiments. Some of the highest results are omitted in the mean 

 given below, but as it is difficult to know what to allow for purely 

 sporadic radium the mean is not very certain. In the case of a spe- 

 cially high result I asked Prof. Emil Werner to determine the ura- 

 nium ; my result was confirmed. My list of mean results on igneous 

 rocks up to the present is the following: 



Basalts (14) « 5. 



Granites (6) 4. 1 



Syenites (1) 6. 8 



Lev/isian gneiss (3) 5.7 



Simplon (32) 7.6 



St. Gotbard (32) 5.1 



The general mean is 6.1. 



From the igneous rocks have originated the sediments after a toll 

 of dissolved substances has been paid to the ocean. It does not of 

 course follow necessarily that the percentage of radium, or more cor- 

 rectly of uranium, in the sedimentary rocks should be less than in the 

 igneous. The residual materials might keep the original percentage 

 of the parent rock, or even improve upon it. There are reasons for 

 believing, however, that there would be a diminution. 



Those sedimentary rocks which have been derived from material 

 formerly in solution offer a different problem. In their case there is 

 little or none of the original materials carried into the secondary rock, 

 and the radio-activity will depend mainly upon how far uranium is 

 precipitated or abstracted with the rock-making substances; in other 

 w^ords, upon how far the waters of the ocean will restore to the rocks 

 what it has borrowed from them. 



This brings me to consider the condition of the ocean as prepara- 

 tory to quoting experiments on the sediments. 



The ocean and its sediments. 



The waters of the ocean, covering five-sevenths of the earth's sur- 

 face to a mean depth of 3.8 kilometers, represent the most abundant 



"This number is to be multiplied by 10"^^, and represents million milliontbs 

 of a gram of radium per gram of material investigated. Throughout the rest of 

 my address this understanding holds, unless where a different meaning is 

 specified. The numbers in parentheses signify the number of different specimens 

 investigated. 



