214 The American Geologist. March, 189-1- 
in shales, together with some of the silicon. The remainder of the sili- 
con was mostly laid down in sandstones. The calcium and magnesium 
are in the limestones. The sodium is in the ocean. The iron is not 
concentrated in any one place, but is diffused generally through nearly 
all the sedimentary rocks. Trie potassium is not easy to rind ; it passed 
into the ocean, and some of it occurs in glauconite and in the terrige- 
nous deposits of blue mud dredged and mapped in the soundings of the 
Challenger expedition. Probably it exists in shales, but we need more 
analyses to determine this. The shales are the residuum after the 
sandstones and limestones have been segregated. A computation of 
the relative amounts of sandstones, shales and limestones, based upon 
the total mean of the analyses of the crystalline rocks, is as follows: 
Table of proportions of the sedimentary rocks. 
Sandstones, 39 per cent., or about two-fifths. 
Shales, 42 per cent., or again about two-fifths. 
Limestones, 19 per cent., or about one-fifth. 
The sodium in the ocean is a definite amount and equals a layer of 
sodium 350 feet deep on the land, which corresponds to one mile of 
crystalline rock. Prom this the author concludes that somewhat more 
than a mile in thickness of crystalline rocks upon areas equal to all the 
present land of the globe must have been worked over to give our sedi- 
mentary rocks. 
37. Volcanite, a,n anortJwclase augite rock chemically like the 
'Incites. William H. Hobbs, Madison, Wis. (Read by Prof. G. H. 
Williams.) A petrographic study of the "bread crust bombs" from the 
eruption of Volcano in 1888. The rock of these projectiles is mineral- 
ogically a trachyte, with anorthoclase replacing sanidine. The struc- 
ture is andesitic and chemically the rock is a dacite. It contains 
interesting magmatic pseudomorphs after augite phenocrysts. 
38. Further notes on the occurrence of Albert ite in New Brunswick, 
Canada. H. P. H. Brumell, Ottawa, Canada. (Read by title.) De- 
scription of several interesting specimens illustrating peculiar condi- 
tions of the mineral in rocks at the base of the Albert series, Albert 
county, N. B. 
39. Alterations of silicates in gneiss at Worcester, Mass. Homer 
T. Puller, Worcester, Mass. In a quarry of gneiss in which orthoclase 
largely predominates are found veins and geodic cavities filled with 
calcite, stilbite, prehnite, and chabazite. The process of formation of 
the last three was described and illustrated in this paper. Bosses of 
hornblende and biotite are also found partly changed to chlorite, and 
this alteration was explained. 
40. Pre-paleozoic decay of crystalline rocks north of lake Huron. 
Robert Bell, Ottawa, Canada. An eroded surface of granite occurs 
under limestone of the Black River formation in the North Channel of 
lake Huron. Rounded pits in the surface of the granite are filled wdth 
fossiliferous limestone and sandstone. Caverns, called "ovens," indent 
steep slopes and bluffs of the granite. Old channels along joints in 
granite and other rocks were deepened by glaciation. Rough, unglaci- 
ated surfaces of Huronian quartzite are shown to have been formerly 
covered by Trenton limestone. These features, and also decay in 
ancient geological times of greenstone dikes cutting the Huronian and 
Laurentian rocks, and the disposal of the detritus from the old crystal- 
line surface, were treated in this paper and illustrated by photographs. 
41. Gabbros on the western shore of lake Champlain. James F. 
Kemp, New York City. The paper opened with a brief statement of 
the general geological relations in the area described. Although in 
many respects obscure, disguised by metamorphism, and concealed by 
drift or forest growth, the succession is probably the following, from 
earliest to latest: a series of gneisses, composed of orthoclase (mostly 
microperthitic), quartz, and mica or hornblende or augite; a series con- 
sisting of black hornblende and pyroxenic schists or gneisses, of crys- 
