Vol.70.] LAVAS A I NORTH llkAi). UT.UiO il \KJJUl K. -AOl 



increase in the basic constituents. There are, of course, other 

 instances in which appeal has been made to the probability of 



fresh accession of magma to the reservoir. The best known is 

 perhaps that of the Berkeley Hills in California, where live or six 

 such recurrences are suggested. In tin's case, however, there is not 

 so clear a succession of lava-flows, and the succession does not 

 include alkaline rocks. 1 



The mineralogica] and textural character of the rocks also supports 

 this idea. The kaiwekite and trachydolerites contain phenocrysts 

 of certain minerals that are not found in the other rocks, which 

 it is here suggested are derivatives from the magma that soli- 

 dified into the kaiwekite. Of these barkevikite, anorthoclase, 

 and the violet augite are the most prominent. The olivine in the-'' 

 rocks also is far less fresh than it is in the basalts, and is often 

 much resorbed. In fact, none of the principal minerals in the 

 kaiwekite (and to a less extent in the trachydolerites) occur in the 

 rocks which are supposed to be derived from the same material. 

 At first sight, these considerations seemed to be an absolute barrier 

 to the idea of the derivation of these lavas from a common source 

 by a process of differentiation. Further consideration, however, 

 somewhat lessened the difficulties. The olivine and perhaps the 

 brown augite-phenocrysts in the basalts might be directly derived 

 from those minerals that were already crystallized in the magma : 

 for, when it came to rest, they would gradually sink to the bottom. 

 because of their relatively high specific gravity, and this fact might 

 well account for their absence in the phonolite anil their presence in 

 the basaltic derivative from that magma. The total absence of the 

 barkevikite and partial absence of the anorthoclase in all the suc- 

 ceeding rocks must obviously involve considerations of a different 

 nature. 



Microscopic study shows that the barkevikite of the kaiwekite 

 and trachydolerites has undergone much resorption, and that the 

 anorthoclase has been greatly corroded. This at least suggests that, 

 under the conditions of eruption of the magma, crystals of these 

 two minerals were unstable. They are certainly crystals of intra- 

 telluric origin formed under conditions that obtain at great depth, 

 but not proper to the composition and other conditions of this 

 magma when it had reached the level of the supposed intermediate 

 reservoir. It appears to be generally recognized at the present time 

 that pyrogenetic hornblende M'ill not crystallize unless the magma 

 contains water or hvdroxvl. Thus Prof. .1. P. Iddings says 



• according to Penfield hydroxyl enters into the composition of hornblende. - 

 . . . The presence of hydrogen in the compound [pyrogenetic hornblende . 

 though in smaller amount than in mica, indicates a similar origin for amphi- 

 bole and mica compounds, namely a hydrolitic one." :< 



1 A. C. Lawson & C. Palache. Univ. Calif. Bull. Dcp. Geol. vol. ii. No. 12 

 (1902). 



2 • Problems of Petrology " Proc. Am. Phil. Soc. vol. 1 (1911) p. 291. 

 a ' Igneous Rocks " New York, vol. i (1909) p. 139. 



