516 GEOPHYSICS 



metallic content from a common deep-seated source. Here, as in 

 several other connections, Professor Suess's theory of " juvenile 

 waters " is very suggestive. It is generally held that many of the 

 great iron deposits are due to magmatic separation. Deposition of 

 lead ores by replacement of calcite is a known process, but takes 

 place under unknown conditions. In some cases replacement of rock 

 by ores appears to me to be alleged without sufficient proof. Pseudo- 

 morphoss is the only adequate test of replacement. 



Erosion appears to be a subject which is capable of more exact 

 treatment than it has received. Weathering and abrasion proceed 

 with a rapidity which increases with the surface exposed per unit of 

 volume. 1 Hence these processes lead to minimum surfaces. Therefore 

 also the mathematics of erosion is essentially identical with that of 

 capillarity. 



Geological climates are as interesting to astrophysicists as to 

 meteorologists and geophysicists. Messrs. Langley and Abbot ap- 

 pear to have evidences of recent variations in solar emanation. If 

 these have been considerable in the course of the period of historical 

 geology, light should be thrown upon them by the paleontology of 

 the tropics. Variations in the composition of the atmosphere must 

 have been very influential in determining both the mean temper- 

 ature of the earth's surface and the distribution of temperature; but 

 so also is the distribution of water. No theory of the glacial period 

 seems generally accepted. Croll's theory is discredited. I have 

 shown to my own satisfaction that the astronomical conditions most 

 favorable to glaciation are high obliquity and low eccentricity of 

 the earth's orbit, 2 but cannot claim any extensive following. If 

 I am right, it should be possible to obtain a definite measure of geo- 

 logical time in years as soon as the astronomers have completed the 

 theory of secular variations in the planetary system so far as to be 

 able to assign the lapse of time between successive recurrences of 

 low eccentricity and high obliquity. 



A most interesting observation, which promises much light on the 

 past history of the globe, is that lavas and strata indurated by lavas 

 retain the polarity characteristic of the locality in which they cooled. 3 

 The time may come when this will lead to determinations of the rela- 

 tive age of lavas, the duration of periods of eruption, and possibly 

 even absolute determinations of date. 



Geology has long, and with some justice, labored under the reproach 

 of inexactitude. As has been illustrated in the preceding pages, the 

 science is still in the qualitative stage, and almost wholly lacks the 

 precision of astronomy. Even its most ardent students have seldom 



1 U. S. Geological Survey, mon. xm, 1888, p. 68. 



2 American Journal of Science, vol. XLVIH 1894, p. 95. 



8 Brunhes and David, Comptes Rendus, vol. 133, 1901, p. 153. 



