TOURMALINE-BEARING ROCKS, MT. HEEMSKIRK. 155 



more abundant, and forms macroscopic bundles of radiating 

 needles, and even granular nodules of quartz and tourma- 

 line. This great increase in the amount of tourmaline 

 present is, vve believe, to be accounted for by some process 

 of differentiation or segregation in the original granite 

 magma, while the presence of tourmaline in separate 

 bundles and nodules may be accounted for by a further 

 jDrocess of magmatic segregation from the already 

 differentiated tourmaline-bearing magma. The tourma- 

 line-aplite magma may have separated from the 

 tourmaline-granite magma in the same wTiy as the tourma- 

 line'-granite magma separated from the normal granite 

 magma. The presence of irregular masses of tourmaline- 

 gianite in the normal granite, and of tourmaline-aplite in 

 tourmaline-granite, is we think, convincing evidence that 

 the three types of rock have originated from one and the 

 same magma, and that the differentiation has taken place 

 before, and not after, the consolidation of the magma. A 

 definite order of consolidation of the three rocks may be 

 observed. Dykes of tourmaline-granite occur in the 

 normal granite, but not in the tourmaline-aplite, while 

 dykes of tourmaline-aplite occur both in the tourmaline- 

 granite and in the normal granite. It is evident therefore 

 that the order of consolidation Vvas (1) normal granite, 

 (2) tourmaline-granite, (3) tourmialine-aplite. 



The composition of the quartz-tourmaline nodules at 

 once suggests that the quartz.-tourmaline reefs are in some 

 way connected with these. In many places the nodules 

 appear to have a tendency to come together and unite to 

 form larger masses of quartz-tourmaline rock. At greater 

 depths, where the cooling of the magma would proceed more 

 slov/ly, this might well take place, and the quartz-tourma- 

 line magma thus formed might be the origin of the quartz- 

 tourmaline dykes. The presence of the veins, however, 

 suggests that a further separation first takes place within 

 the quartz-tourmaline magma, namely, a separation into a 

 highly aqueoii:; and a less aqueous magma. The former 

 would be virtually a saturated solution, and in this would 

 be concentrated the heavy metals originally contained in 

 the magma, since, from the common occurrence of these 

 metals in mineral veins, we know them to be specially 

 soluble in heated waters. Assuming, nov/, that such masses 

 of quartz-tourmaline magma were intersected by fissures 

 formed by the contraction of the granite in cooling, the 

 more highly aqueous portion, being the more fluid, would 

 be first erupted, and ov^ing to its highly-heated condition 

 and the presence of such powerful mineralising agents as 



