Geology and Mineralogy. 71 



modern geologists are accustomed to use this hypothesis in 

 explaining the distribution of fossil fauna and flora. Mr. 

 Harmer shows, on the other hand, that wide-spread climatic 

 changes, sufficient to determine the physical and biologic char- 

 acter of a region, may be produced by variations in the relative 

 position of areas of high and low barometric pressure. The 

 former humid condition of the Sahara and of the Great Basin, 

 the existence of the Mammoth on the shores of the Polar Sea, 

 the non-glaciated Alaskan regions, and the secular variations of the 

 European and American ice-sheets, are factors which have aided 

 the author in determining the areas of low and high barometer 

 and the prevalent storm-tracks of the Pleistocene Epoch. Many 

 meteorological difficulties are avoided by adopting the hypothesis 

 that the maximum glaciation of the eastern and western conti- 

 nents was not contemporaneous. While giving due weight to 

 the older theories of the cause of the Glacial Age as well as to 

 the more recent theories of Prof. Chamberlin and Dr. Ekholm, 

 Mr. Harmer is "inclined to think that the minor variations of the 

 Pleistocene, the pre-historic, and the historic periods may have 

 belonged to one great series of events, and have been alike due to 

 the cause which gives Great Britain its variable seasons at the 

 present day, namely, to alterations in the directions of the preva- 

 lent winds." h. e. g. 



6. The Relative Density of Fluid and Solid Magmas. — The 

 question as to whether contraction or expansion takes place when 

 an igneous rock is formed by the solidification of a fluid magma, 

 is one that had been often discussed in connection with volcanic 

 theories. It is generally assumed that the former is the case, and 

 the experiments of Barus, detailed in this Journal ((2), xlii, 498, 

 1891), are conclusive on the point. Further experiments have 

 been recently carried on by C. Doelter. He shows that in all 

 the cases investigated by him, as given in the table below, the 

 density of the molten fluid is less than that of the corresponding 

 solid rock ; there being a difference of O^-O'S in most cases. 

 In order to obtain accurate results, he remarks that it is necessary 

 to have the molten mass at a temperature considerably above that 

 of fusion, so that its fluidity may be complete. In the case of 

 melanite (lime-iron garnet), for example, which fuses a little 

 above 900°, the experiment w T as tried between 1050 and 1100°. 

 Instead of using a fragment of the same material in testing, a 

 number of different minerals of appropriate densities, and with 

 fusing points considerably above the temperature in question, 

 were adopted. By noting which of these sank and which of 

 them floated in the fluid, the density of the fluid was obtained 

 with a fair degree of accuracy. The accompanying table gives 

 the result for two minerals and five basic rocks. It will be seen 

 that the density of the solidified material in the glassy form is 

 considerably less than that obtained when solidification takes 

 place slowly and the product obtained is more or less crystalline. 

 Furthermore, the last density approximates towards the original 



