W. N. Benson — Origin of Serpentine. 721 



is also adopted by Graham (1917), who argues that the 

 solutions were not carbonated but merely silicic, since 

 there is a noteworthy absence of any carbonates, and it 

 seems improbable that an amount of carbonate in bulk 

 from a quarter to a twelfth of the volume of serpentine 

 could have been removed in solution. Dr. Bell thinks 

 that the hydration of the dunite in the Parapara District, 

 N. Z., may be the result of thermal waters accompanying 

 the acid intrusive rocks which are associated with the 

 ultrabasic rocks (Bell 1907), but there is no suggestion 

 of the presence of such rocks in the quite similar serpen- 

 tines of the Hokitika District (Bell 1906). Otis Smith 

 (1904) has suggested that the serpentinization of the 

 peridotites in the Mt. Stuart district, "Washington, 

 U. S. A., was effected by magmatic waters emitted from 

 the mass of granodiorite by which they have been 

 invaded, but again the presence of such invading masses 

 is by no means a constant feature in the occurrences of 

 serpentine along the Pacific slopes of the United States. 

 The continuous zone of serpentine between the dunite 

 and the pyroxenite of the Tagil complex in the Ural 

 Mountains (see fig. 4), as shown in Wyssotzky's care- 

 ful map (1913), may illustrate another instance of this 

 mode of formation, though it must be noted that here it 

 has been suggested that the pyroxenite is a differentiate 

 formed in situ, and simultaneously with the dunite, 

 rather than a latter intrusion. The same suggestion has 

 been made in Tasmania, by Waterhouse (1916) in the case 

 of the Heemskirk District, where there have been both 

 gabbro and granite intrusions following the peridotite in 

 the one igneous epoch, and it has also been urged by Mr. 

 Twelvetrees (1917) in explanation of the serpentines of 

 Anderson's Creek, which have been invaded by granite. 

 In the last two cases there is a large development of 

 granite with the serpentine, but in others as Graham 

 remarks (1917) "it may be objected that the number and 

 size of the exposed granite dykes and masses are totally 

 inadequate to have been responsible for serpentinization 

 on the scale which has actually occurred"; but he sets 

 the objection aside as based merely on a matter of opin- 

 ion as to the amount of water that might accompany a 

 granite-intrusion. There is also to be considered the 

 possibility of the occurrence of large unexposed masses 

 of granite where but few veins are visible. In the Great 



