Smith. — Alkaline and Nepheline Bocks, Westland. 135 



inclusions ; and fluid-pores, so typical of the quartz of granite 

 porphyries, are absent. A few minute glass inclusions without 

 bubbles and the riebeckite inclusions already mentioned are all 

 that can be made out. The perthite phenocrysts show less 

 signs of corrosion and a greater variety of shape. Some are 

 rectangidar to square plates ; others of elongated blade-like 

 habit, occasionally reaching 7 mm. in length by 0-4 mm. in 

 breadth. Those with elongated habit have undulose extinction. 

 In many of the sections of the mineral both the component 

 feldspars extinguish simultaneously ; in others there is a dis- 

 tinct angular interval between extinction ; and there is no 

 crystalline continuity between the feldspars of the phenocrysts 

 and the sectors of spherules radiating from them. In some of 

 the perthites the feldspars are orthoclase and anorthoclase ; 

 in others the nature of the triclinic feldspar has not been 

 determined. In some of the slices there is an occasional 

 crystal of sphene. Omitting the rare accessory sphene and 

 the iron-ores which may possibly be of secondary origin, 

 the sequence of crystallization in the consolidation of the 

 rock is: (1) riebeckite; (2) quartz and perthite phenocrysts; 

 (3) spherulitic growths ; (4) the quartz mosaic and micro- 

 graphic intergrowths representing the balance of the ground- 

 mass. It is just within the bounds of possibility that the 

 riebeckite may occur in two generations, that in the interstitial 

 groundmass being the prior, and that in the spherules a subse- 

 quent crystallization contemporary with the growth of the 

 spherules ; but the evidence is strongly in favour of the amphi- 

 bole having conformed to the normal order of consolidation, 

 and only occurring in one generation. The development of the 

 spherules appears to have proceeded in the following manner : 

 Starting from a centre, the nature of which cannot be dis- 

 tinguished, the formation and consolidation of the cryptopeg- 

 matite proceeded outwards, the process of crystallization push- 

 ing the larger and broader plates of riebeckite before it, but 

 turning the needle-like blades — which it ultimately enclosed — 

 in the direction in which they offered the least resistance to the 

 process — that is, radially. Needles of riebeckite which escaped 

 this first capture were seldom included in the outer zone of the 

 spherule. Any mass too large to be moved by the energy of 

 crystallization was included in the radiating body of the 

 spherule, the growth of which proceeded on the further side of 

 the inclusion as if the obstacle did not exist (see Plate XVIII, 

 fig. 4. where a grain of quartz is included between the centre 

 and the periphery of a spherule). After the process had con- 

 tinued outward beyond the length of the included needles of 

 riebeckite, the micrographic growth of the spherule is com- 



