JF. S. Greslei/ — A Coal Pebble in a Coal-ncam. 157 



leucite syenites on the one hand and the essexite group on the other. 

 In this position it stands rather alone, its three nearest neighbours 

 being ijolite from Umptek and augite-syenites from Hedrum 

 (Norway) and from Yogo Peak. 



The dyke-rock falls nearest to an ijolite from liwaara (Finland), 

 leucite-syenite from Shonkin Creek, and amphibole-malignite from 

 Poobah Lake, Canada. 



Borolanite is more acid and more alkaline than either the nepheline- 

 syenite, leucite-syenite, or leucite-tinguaite-porphj-ry of Magnet Cove. 



With reference to the Christiania rocks, the alkali and silica ratios 

 of borolanite are about half-way between those of laurdalite and the 

 essexite of Solvsberg. Its alkali ratio is nearly identical with that 

 of laurvikite from Tonsberg. 



In a recent paper * Dr. F. H. Hatch has published a new scheme 

 of classification for the plutonic rocks. In this the basic rocks are 

 divided into the alkali-gabbro and gabbro families ; the alkali-gabbros 

 are further divided into nepheline-gabbro, essexite, shonkinite, and 

 kentallenite sub-families. The rock of Camas Eilean Ghlais falls in 

 the nepheline-gabbro sub-family along with theralite and covite. 

 Borolanite itself appears to fall in the shonkinite sub-family, as it 

 iindoubtedly belongs to the potash series. By plotting the ' rock- 

 pictures ' by Brogger's modification of Michel-Levy's method it can 

 readily be seen that borolanite bears affinities to the shonkinite of 

 Poobah Lake, and also that it belongs to the same series as plauenite 

 (potash-syenite of Plauen, Dresden). 



An isolated analysis of one specitnen from such an interesting 

 locality is of very little use in determining the aflS.nities of the rock, 

 and it is hoped that further analyses of the Cnoc-na-Sroine complex 

 will be made, so that it may be possible to compare these rocks with 

 those of Magnet Cove and Christiania, and to arrive at some definite 

 conclusions as to the origin of borolanite. 



In conclusion, I have to acknowledge my great indebtedness to 

 Mr. Bastall for his very valuable assistance, and also to Dr. Shand for 

 useful information and suggestions. The analysis was made in the 

 Mineralogical Laboratory at Cambridge, and I have to thank 

 Dr. A. Hutchinson for his assistance and advice in this part of 

 the work. 



IV. — Coal Pebble in a Coal-seam. 



By W. S. Gresley, Assoc. M. lust.C. E., F.G.S. 



Recorded Coal Pelhles. 



LOGrAN, in 1840, noted the occurrence of pebbles of cannel in 

 the roof of a coal-bed in South "Wales ; also of pebbles of 

 coal in the * Pennant ' rock (Proc. Geol. Soc, vol. iii, p. 276). 

 De la Beche refers to coal pebbles in the Monmouth and Glamorgan- 

 shire Coalfield,- and H. K. Jordan to similar forms of coal in the 



^ Science Progress, No. 10, October, 1908. 



^ [Mem. Geol. Survey, 1846, vol. i, pp. 159, 193, etc. ; see also Geology of the 

 S. Wales Coal-field, 1907, pt. vii, by A. Strahan and others. — Ed.] 



