Gcroser 4, 1918] 
evidence in favor of the view that regional 
dolomites are of shallow water origin. The 
bearing of this conclusion upon the coral reef 
problem is made clear from the following quo- 
tation :3 
The author regards the evidence of dolomitiza- 
tion of fringing reefs of coral islands, the occur- 
rence of dolomite immediately below phosphate 
beds, on the hill tops of Christmas Island, the rise 
in the magnesian content of the limestones of the 
Funafuti bore between 15 and 30 feet, as definite 
and strong evidence of the shallow water origin 
of dolomite in coral islands. It is claimed that 
this view is consistent with the chemical evidence 
quoted above of the reversal of the solubilities of 
calcium and magnesium carbonates in carbonated 
water between the pressures of one and four at- 
mospheres. In addition, the evidence, cited above 
from more ancient dolomites showing their inti- 
mate associations with independent evidence of 
shallow water such as fossils, current bedding, 
conglomerates, and oolitie structures, is so consist- 
ent and so in accordance with the evidence from 
modern coral limestones, that the author takes the 
view that wherever a ‘‘contemporaneous’’ or re- 
gional dolomite is found to oceur, it may be re- 
garded as having originated in shallow water. If 
this be granted, it follows that such upraised coral 
islands, like Ngillangillah, now 510 feet high, and 
Vatu Vara, now 1,030 feet high, which are dolomi- 
tized from top to bottom, must have originally 
been formed of shallow water limestones accumu- 
lated by subsidence to at least 500 to 1,000 feet 
respectively before elevation set in. The atoll of 
Funafuti whose surface is practically at sea level 
must have been built up of shallow water lime- 
stones accumulated during subsidence, which must 
have amounted to about 1,100 feet at least since 
the cores from 635 feet to 1,114 feet consist en- 
tirely of limestones which have passed through 
the process of dolomitization. 
In the writer’s experience with regional dolo- 
mites of undoubted secondary origin, he has 
encountered considerable evidence in support 
of the contention that many of them represent 
shallow water deposits, but he is not yet pre- 
pared to conclude that all replacement dolo- 
mites are of this origin. The most striking 
evidence bearing on this question that has 
come to the writer’s attention, has been ob- 
tained in connection with his study of the 
8 Ibid., p. 200. 
SCIENCE 
dol 
limestones of the Osage and Meramec series, 
of Mississippian age, in the Mississippi Valley. 
In Ste Genevieve county of southeastern 
Missouri, these limestones, with one exception, 
possess all the ear-marks of clear, open sea de- 
posits (see table). They attain their maxi- 
TABLE 
Thickness 
Serles Name of Formation in Feet 
Saint Louis limestone ....... 150 
Meramec i 
Spergen limestone .......... 160 
Warsaw formation .......... 150 
Osage....2 Keokuk limestone .......... 30-40 
Burlington limestone ........ 75 
mum development there, are all, with the ex- 
ception of the Warsaw, unusually pure, and, 
barring a small break of local significance at 
the base of the Warsaw, are conformable. The 
paucity of dolomite in this thick series of lime- 
stones isremarkable. With the exception of an 
impure bed of dolomitic limestone in the upper 
portion of the Warsaw, which may well be of 
clastic origin, and a thin, imperfectly dolo- 
mitized layer in the Saint Louis at the station 
of Little Rock, no dolomite was observed dur- 
ing a careful study of the whole section. 
In southeastern Iowa and adjacent portions 
of Illinois, on the other hand, very different 
conditions are met with. In this region all 
the formations show indications of having 
been deposited in shallow, oscillating seas, the 
evidence being most pronounced in the 
Spergen and Saint Louis limestones, and dolo- 
mite is a very important constituent of every 
member of the series. 
The Burlington limestone maintains ap- 
proximately the same thickness here as in 
southeastern Missouri, but beds of brownish, 
impure dolomite, some of which pass locally 
into shale, are interbedded with the limestone 
and constitute more than fifty per cent. of the 
formation. 
The Keokuk consists of interbedded layers 
of shale and limestone, some of the latter be- 
ing dolomitized locally; and the Warsaw is 
made up chiefly of argillaceous shale but bears 
occasional lenticular beds of limestone, some 
of which are imperfectly dolomitized. 
The Spergen limestone of southeastern Iowa 
