962 A TREATISE ON METAMORPHISM. 



ought to correspond with the percentage found in the original igneous rocks. 

 Therefore the discrepancy above noted indicates that there are factors left 

 out of and errors in the above equation. In reference to the similar equations 

 of most of the other elements it will be seen that the determination of the 

 amount of the element as made from the sediments usually shows a defi- 

 ciency as compared with the original rock. It will be explained that one of 

 the reasons for this is that determinations of the average composition of 

 the sedimentary rocks based upon analyses of the shales, sandstones, and 

 limestones neglect the particular formations in which an element is segre- 

 gated to the largest extent. In the case of silica the reverse is the case. 

 The sandstones in which silica is concentrated are included; whereas other 

 exceptional compounds in which there is depletion of silica are not included. 

 Consequently we should expect in the case of this compound that the esti- 

 mates made upon the basis of the sediments should be in excess of the 

 amount found in the original igneous rocks. While the discrepancy is thus 

 parti}' explained, I suspect that it is due more largely to an error in the 

 determination of the amount of silica in the shales. Few analyses have 

 been made of this class of compounds. It will be seen on subsequent 

 pages (see pp. 984-986) that alumina shows a discrepancy just the reverse 

 of that of silica. It appears to me probable that when we have better rep- 

 resentative analyses of the shales, taking into account their composition 

 and masses, the silica will be lower and the alumina higher than in the 

 analyses given. 



CARBON. 

 AMOUNT OF CARBON. 



In 1891 Clarke estimated the carbon as 0.21 per cent of the lithosphere, 

 hydrosphere, and atmosphere together. This corresponds to 0.22 per cent 

 of the original rocks, 0.002 per cent of the hydrosphere, and only .0127 

 per cent by weight of the atmosphere. In 1900 his estimate for the 

 lithosphere was reduced to 0.12 per cent. Recently Letts and Blake, ffl as a 

 result of careful consideration of all the data and new work, conclude that 

 carbon dioxide is present in the atmosphere to the amount of three volumes 

 in 10,000, which corresponds to about 0.045 per cent by weight. 6 



"Letts, E. A., and Blake, R. F., The carbonic anhydride of the atmosphere: Sci. Proc. Eoyal 

 Dublin Society, vol. 9, new series, pt. 2, 1900, p. 172. 



t> Mendeleeff, D., The principles of chemistry, translated by George Kamensky: Longmans, Green 

 & Co., London, 6th ed., 1897, vol. 1, p. 238. 



