80S im;of. i\ maeshall on the sequence or [Dec. 1 9 14- 



every connecting-link between the kaiwekites and the trachy- 



clolerites : that is. between the volcanic representatives of the 

 kurvikites and the essexites respectively. 



Elsewhere along tins south -eastern coast of Xew Zealand volcanic 

 rocks in many places traverse the Tertiary rocks and the schists. 

 All of these appear to be basalts of a more or less uniform type, 

 and similar to the basalts of the Dunedin district. At Kakanui, 

 SO miles distant from Dunedin, there is a coarse submarine tuff, 

 which contains large crystals of hornblende, and an acid oligoclase, 

 as Avell as other minerals and fragments of peridotites. These have 

 been described by Dr. J. Allan Thomson. 1 



V. Origin of the Different Lavas. 



A consideration of the chemical and mineralogical characters 

 of this sequence of lavas, as described, suggests that at least four 

 different explanations might be advanced to explain its features : — 



(1) A separate magma-tic origin might he assigned to each of the lavas. 



(2) Two magmas might he supposed to exist side hy side, and their zone 



of contact might give rise to the mixed type. 



(3) Attempts might be made to explain the whole series as a result of the 



differentiation of one parent magma. The different types would 

 then be explained on the hypothesis that effusion would take place 

 from different parts of the magma as it gradually cooled and various 

 parts of it acquired the composition that characterizes the lavas 

 which issued from it. 



(4) The intermediate types, that is. the kaiwekite and the trachydolerites. 



may be regarded as undifferentiated portions of the original magma 

 with which the volcanic reservoir was supplied on three occasions 

 at least. From this original magma the other types may have 

 been derived by some process of differentiation. 



(1) It is hardly likely, at this stage of the progress of petro- 

 graphical science, that any geologist would be found to support the 

 idea of a separate magmatic origin for each type of lava repre- 

 sented in the series. Although in the island of Hawaii the two 

 volcanoes, Kilauea and Mauna Loa, are quite independent one of 

 the other in their periods of activity, the products of this activity 

 are similar. 



In the present case, however, it would be necessary to suppose 1 

 that at least four volcanoes existed in such proximity that, in their 

 various periods of activity, their lava-flows would have overlapped. 

 As stated earlier, the field-evidence shows that the lavas all flowed 

 down the same slope, and that the time-interval which separated 

 different eruptions was in no case great. There is. in addition, 

 so evident and close a relationship, from a chemical standpoint, 

 between the various members of the sequence as to necessitate some 

 amount of community of origin. 



(2) The existence of two primary magmas and their combination 

 in different proportions to form the multiplicity of igneous rock- 



1 Trans. N.Z. Inst. vol. xxxviii (1905) pp. 482-05 ; also ' Inclusions in some 

 Volcanic Eocks ' Geol. Mag. dec. 5, vol. iv (1907) p. 497. 



