ROOFING SLATES OF NEW YORK AND VERMONT. 39 
The significance of these analyses and microscopic examinations is that this slate 
n weathering first loses its carbonate, then its carbonaceous matter, then its pyrite, 
md has its feldspathic material, if any, changed to kaolin, while its muscovite 
emains partly unaltered and its quartz largely so. The bleaching process in the 
lates of Lehigh and Northampton counties, Pa., apparently proceeds by these steps: 
1. Removal of carbonates of lime and magnesia, accompanied by oxidation of the 
errous carbonate and deposition of limonite. 
2. Oxidation and removal of carbon; also oxidation and removal of pyrite. 
3. Removal of almost all limonite. 
In general, therefore, while slate in weathering returns to a clay, this clay, in some 
:ases, differs from that in which the slate must have originated by the absence of 
:arbon and the small amounts of silica, potash, and soda, and in others by the 
ibsence or extremely low percentages of carbon, lime, and iron. 
<TOTES ON THE CHEMICAL COMPOSITION OF THE ROOFING SLATES 
OF EASTERN NEW YORK AND WESTERN VERMONT." 
By W. F. Hillebrand. 
Before the analyses of any of the slates were undertaken, it was well understood 
hat the chief question of economic importance to be solved was the cause of fading 
>bserved in some, but not all, of the green slates after longer or shorter exposure to 
atmospheric influences. Attention was therefore primarily directed toward its eluci- 
lation, and the evidence accumulated will be herein set forth. Naturally, in the 
ourseof the work, opportunity was afforded for drawing conclusions as tothecombi- 
Lation of certain chemical elements concerning which the microscope could furnish 
10 satisfaction. A few remarks in this connection may therefore well be in place. 
It soon became clear that there was no apparent connection between the percent- 
ge of iron sulphide (FeS 2 ) in any green slate and its liability to fade; that, further- 
Qore, the visible sulphide was of an extremely resistant type, and often after years 
i exposure retained its original luster undimmed; that it was probably true pyrite. 
?he possibility remained that there might be finely divided FeS 2 in the much more 
eadily decomposable form of marcasite. Microscopical evidence made it clear that 
f present at all it was in amount altogether too trifling to produce the fading effect 
ften observed. 
It was found that the ferrous iron, other than that in the pyrite, was almost entirely 
n a very soluble condition. Even when carbonates were absent, a large proportion 
f it dissolved in moderately dilute acids after very few minutes boiling. But the fer- 
ous iron thus shown to exist in the silicate minerals could not be accepted as the 
ause of fading, because it exists in all slates; and some of the most ferruginous, as 
14/, are ''unfading." Microscopical evidence is also opposed to oxidation of ferrous 
ilicates as the cause of fading. 
There remained as its sole probable cause the carbonate found in nearly all of the 
lates. Its chemical behavior toward acids proves beyond question that this carbo- 
;ate can be in no observed case calcite. Were it even in small part so, an instanta- 
leous disengagement of gas would be apparent on adding acids, whereas their action 
3 tardy and becomes pronounced only on application of, heat. To those familiar with 
he behavior of different carbonates this evidence alone is decisive. 
The carbonate is, therefore, an isomorphous mixture, and indeed the C0 2 found, as 
ompared with the CaO, shows at once a great excess of the former. That CaC0 3 is 
ne of the chief isomorphous constituents appears from the fact that absence of C0 2 
niformly accompanies the lowest trifling percentages of CaO, and that increase of 
)aO carries with it an approximately corresponding one of C0 2 . The C0 2 is usually 
n molecular excess of the CaO, and sometimes of the MnO and FeO as well (compare 
a Reprinted from Nineteenth Ann. Rept., part 3. The analyses and microscopic descriptions here 
3ferred to will be found under New York and Vermont slates. 
