246 Prof. Horsford on the Rocks of the Florida Reefs. 
length, according to the proportions . the ingredients. The 
processes are both chemical and mechanical. 
4. Gypsum from which the two Ret of water of erystalliza- 
tion have been expelled by heat, rapidly hardens upon being mixed 
with water. This is ascribed to the reunion of the sulphate of 
lime with the water. 
Do either of the above processes suggest the method by which 
the rocks of the Florida reefs have been hardened ? 
e facts presented in the furnished specimens are as follows: 
The rock formed under water exclusively is composed of grains 
of size less than that of a mustard seed, which, to the naked eye, 
appear quite i Sagar and of uniform diameter. More carefully 
examined with a microscope, they are found to be far from regu- 
lar in form or a in size, but present numerous depressions 
and prominences. Distributed throughout the intervening spaces 
is a fine deposit of carbonate of lime, which adheres with consid- 
erable tenacity to the surface upon which it rests, 
he surface or crust-rock, though not strictly homogeneous, 1s 
composed of particles so minute as not to be distinguished from 
each other. ny dissolves in hydrochloric acid, leaving a floceu- 
lent residue. The solution, when evaporated to. dryness and 
ignited, readily redissolves in hydrochloric acid, with only an oc 
casional residue. The solution gives no precipitate with chlorid 
of barium. Nitrate of silver gives, in a nitric acid solution, @ 
white precipitate soluble in ammonia. The aqueous extract 
— to alcohol flame the characteristic soda tint. The pes wde - 
of iron, magnesia and silica. The former two were wantl 
most of the specimens examined, and the silica in some. Numer 
ous specimens were examined for alumina, without in avy ine 
stance finding a trace of this substance.* 
In a quantitative analysis by Homer, and another by Mariner, 
the following results were obtained. 
* T examined, also, all the species of coral at m my command, without finding a trace 
of alumina in any of them. The hydrochloric acid oe . of the coral was ee 
tated with ammonia. The | washed precipitate was digested for several hours The fil 
potassa Girovionaly t yand filtes we fer 
trate was then neutr alised with hydrochloric acid, and am ther in 
stag fo for several hours, there appeared filamen ts which ‘were war ofits te # 
ni id, and. w. with the mi : 
eroscope, 
paper; they had been Seived pee the filter. . Beside these, there eireny ais " 
The quantities employed were, in several instances, from a z to 2 
osha of ma al. There were examined, Millepora alcicornis | 
vielen subacrial ein ro} —, ie 
