480 SCIENCE PROGRESS. 



author is thus brought to the conclusion that this feature is 

 not restricted to surface-flows, but he believes that rocks 

 exhibiting it must have been erupted under very nearly 

 surface condition. This is to some extent in accordance 

 with the opinion expressed by Fox and Teall with respect 

 to the dolerite associated with the radiolarian chert of 

 Mullion Island, Cornwall. This rock, " separated into rude 

 rolls by curvilinear joints," they suggest may represent a 

 submarine flow injected between the layers of the stratified 

 series near the sea-bed during deposition. The associa- 

 tion of these curious spheroidal rocks with radiolarian cherts 

 at a number of localities in California and elsewhere seems 

 significant, but the mechanism of igneous eruptions under 

 a great pressure of water is a subject upon which it would 

 not be safe to speculate at present. 



We have already noted Ransome's account of the 

 serpentine of Angel Island, which is demonstrated to be a 

 decomposed diallage-rock. Palache (37) has described the 

 serpentine of the Potrero, San Francisco, which is also of 

 igneous origin and intrusive in the sandstones. In this 

 case the rock has been derived from one consisting of 

 olivine, enstatite, diallage, chromite, and magnetite, and so 

 belonging to the lherzolite type. Lawson finds serpentine 

 rocks in other parts of the district to be in all cases derived 

 from igneous rocks, peridotites or pyroxenites. 



The work embodied in the various memoirs which we 

 have cited has q-one far towards removing some serious 

 difficulties felt by many readers with reference to the 

 geology of the Pacific Slope as interpreted in Becker's 

 monograph on the Quicksilver Deposits (1888). The San 

 Francisco Sandstone and its associated rocks correspond 

 with the Knoxville group of that writer, who endeavoured 

 to explain the characters of many of these rocks by various 

 obscure processes of metamorphism operating upon sedi- 

 mentary deposits. The bedded jaspers are now proved to 

 be not silicified shales, but original siliceous deposits, largely 

 due to radiolaria, though Lawson believes much of the 

 silica to have been chemically precipitated. The "pseudo- 

 diabases " and " pseudodiorites " of Becker, formerly stated 



