W. N. Benson — Origin of Serpentine. 705 



other hand serpentines may occur in completely hydrated 

 condition in the center of deserts (Merrill 1899, Holland 

 1899). For this reason Merrill emphasized the ineffec- 

 tiveness of atmospheric weathering, stating his belief 

 that serpentinization is a deep-seated process due to 

 waters coming from a considerable depth, which may 

 even have been present in the parent magmas at the time 

 of their intrusion. To the first part of this, Bonney has 

 given his partial adherence (1899). Holland, rejecting 

 atmospheric weathering as a cause of serpentinization, 

 gave quite another explanation which is considered 

 below. 



IV. The Source of the Water for Serpentinization. 



With the exclusion of surface water acting at small 

 depths, from among the possible agents of serpentiniza- 

 tion, attention must be directed to the deeper sources 

 of water which may have effected the hydration of the 

 intrusive masses. Four sources may be considered. 



a. Water diffused from the invaded rock in which it was origin- 

 ally contained into the intrusive peridotite. , 



o. Water forming portion of the underground circulation of 

 at least partly epigene or meteoric origin. 



c. Water issuing from the ultrabasic magma itself during the 

 last stage of its consolidation. 



d. Water issuing from magmas intrusive into or near the ultra- 

 basic mass. 



a. Local diffusion of interstitial water. 



The hypothesis of the production of serpentine by the 

 diffusion into peridotite of the water contained in the 

 invaded rock might explain the marginal serpentinization 

 of intrusive masses, if there were to be considered only 

 such masses as are intrusive into comparatively unal- 

 tered sediments, with a high content of water. It is much 

 more difficult to apply it in the case of large masses of 

 serpentine that are intrusive into gneissic rocks such as 

 that in the Geisspf ad Pass, and in the St. Gotthardt Tun- 

 nel; and it is particularly difficult of application when 

 the masses are almost completely serpentinized through- 

 out, as in the case of that in the gneiss at Zoblitz in Sax- 

 ony. In the last mentioned area, the serpentine, in such 

 samples as have been seen by the writer in various 

 collections or obtained by himself, exhibits the normal 



