154 Scientific Proceedings, Royal Dublin Society. 



these lighter shells. "While the olivine in the nodules is associated with a 

 chrorae-spinellid and a little pyroxene, the kelyphitic zone consists of 

 tremolite and magnetite. 



Schrauf's " kelyphite " is now generally regarded as a product of inter- 

 action, and we may probably cite its formation round garnet in peridotite 

 as evidence of a zonal replacement of the garnet. Sir T. H. Holland, 37 on the 

 other hand, considers the micropegmatitic zone round garnets in certain 

 diorites, charnockites, and pyroxenites as a phase in garnet growth, rather 

 than as a siliceous interchange with the more basic matter of the garnet. 



As an example of how zones of interaction may become relict zones, we 

 may cite the concentration of pyroxene and garnet by a process of selective 

 absorption, which has been pointed out in connexion with igneous contacts in 

 Ireland. 38 In both the cases studied, the minerals have arisen as the result of 

 metamorphism in the rock attacked, and have then proved more resistant to 

 absorption than their companions in the altered mass. 



This leads us to enquire whether the production of minerals by contact- 

 metamorphism, which may, of course, work upon a regional scale, does not 

 commonly imply a transference of material to the altered rock, accompanied 

 by a removal of other material into the invader. This outward streaming as 

 the concomitant of an inward streaming clearly concerns the formation of 

 concentric shells round about the nuclei in orbicular rocks. The range of 

 density in the common rock-forming minerals is not sufficient to allow of the 

 utilisation of much imported material within the limits of the block attacked. 

 In a large number of observed cases, moreover, basic inclusions become 

 replaced in part by quartz and felspar. Just as carbon dioxide passes out 

 during the production of silicates in a limestone, so magnesium and iron, at 

 any rate, disappear from a basic inclusion by diffusion outwards. H. J. 

 Johnston-Lavis 33 in 1894 pointed out this exchange of constituents as a 

 cause of variation in igneous rocks, and styled his hypothesis, perhaps some- 

 what vaguely, " osmotic." Endell, 40 when urging the ionic nature of such 

 diffusion, gives very just recognition »to Johnston-Lavis as the pioneer. 

 A. C. Lawson 41 connected zoned orbicular structure with osmotic diffusion in 



37 " On the Origin and Growth of Garnets, and of their Micropegmatitic Intergrowths 

 in Pyroxenic Rocks," Bee. Geol. Surv. India, vol. xxxix (1896), p. 20. 



38 G. A. J. Cole, "On Contact-phenomena at the Junction of Lias and Dolerite at 

 Portrush," Proc. R. Irish Acad., vol. xxvi, Sect. B. (1906), p. 62. 



30 " The Causes of Variation in the Composition of Igneous Rocka," Natural Science, 

 vol. iv (1894), p. 139. 



40 Op. cit. (25), p. 131. 



41 "The Orbicular Gabbro at Debesa, San Diego County, California," Univ. Calif., 

 Bull. Depart. Geol., vol. iii (1904), p. 396. 



