376 J. F. KEMP — GRANITES OF ATLANTIC COAST 



T. Epidotic granite. By L. V. Pirsson. Ainerica.ii Journal of Science, Novem- 

 ber, 1 89.'], page 373. 



II and III. By F. W. Love, formerly Instructor in Chemimtry, Cornell University 

 The analyses were made for this paper some years ago. On account of the acci- 

 dental loss of the solution in which the mixed alkaline chlorides were to have l)een 

 determined, the soda was estimated by difference. The potash was directly 

 weighed. 



IV. By H. T. Vulte, Instructor in Chemistry in Barnard College, Columbia Uni- 

 versity. This analysis and VI were made for this paper. 



V. By L. P. Kinnicut, of the Worcester Polytechnic Institute, for the Red Granite 

 Company, from w^hom it was obtained. 



VI. By H. T. Vulte. 



Specific gravities of samples different from those on which the analyses 

 have been based have been determined as indicated below. Samples of 

 from 100 to 800 grams were employed, and where possible polished spec- 

 imens were used : 



Westerly, gray variety 2.654 



Millstone point 2.660 2.671 



Stony point, Norcross quarry, finer grain 2.626 Coarser, 2.634 



Stony point, Brooklyn quarry 2.640 2.650 



Leets island, gneissoid 2.649 



Stony creek, mica-gneiss 2.704 



Masons island, aplite 2.596 



Masons island, basic inclusion 2.860 



Stony creek, basic inclusion 2.987 



Stony creek, basic inclusion 2.880 



Stony creek, basic inclusion (Brooklj^i quarrj"-) 2.957 



An examination of the analyses brings out the fact that, excei)t for 

 the one from Millstone point, the granites have -quite uniform percent- 

 ages of silica. The ones from Rhode Island range somewhat higher in 

 lime than do those from Connecticut. The magnesia is very low except 

 in the one from Conanicut. In the relations of the alkalies there are 

 some interesting contrasts. In I, IV, and VI the soda is relatively high. 

 In II and III a very decided preponderance of the orthoclase molecule 

 is indicated, and in the sample used for V there must have been a great 

 excess of orthoclase or microcline. The anal3'^sis would indicate one of 

 the purest orthoclase (or microcline) granites of which an}'- record has 

 been made. While samples ma}^ be obtained in the quarries which 



