476 
BULLETIN OF THE BUREAU OF FISHERIES 
Diatoms are at their maxima in this part of the sea in spring. Hence, the 
general correspondence between the silica curve and the fluctuations of the diatoms 
is at least suggestive. Furthermore, only a very slight difference in the concentra- 
tion of silica dissolved in the sea water may be needed to control the multiplication 
of diatoms — perhaps less than one part in two million of water. 
To test whether a similar parallel between seasonal concentration of dissolved 
silica and abundance of diatoms would be found in the Gulf of Maine, samples of 
sea water of about 8 liters each were collected monthly off Gloucester from December 
28, 1920, to October 26, 1921, and shipped in 2-gallon tinned-iron cans to the chemical 
laboratory of the United States Geological Survey for analysis. The determina- 
tions for silica were made by Dr. R. C. Wells, who has described his methods (Wells, 
1922), the results being as follows: 
Soluble silica in sea water collected at the surface about 1 mile south of Eastern Point Light , Gloucester, 
Mass. 
Date of collection 
Silica as 
Si 02 in 
milli- 
grams per 
liter of 
water= 
parts per 
million 
Dec. 28, 1920 
1.5 
Jan. 26,' 1921 
2. 5 
Mar. 2, 1921 _ 
2.9 
Mar. 25, 1921 
1.4 
Apr. 25, 1921 
.3 
May 26, 1921 
.4 
Date of collection 
Silica as 
Si 02 in 
milli- 
grams per 
liter of 
water *= 
parts per 
million 
June 27, 1921 
/ 1 1.9 
l ] .9 
July 27, 1921 
.G 
Aug. 26, 1921 
.3 
Oct. 26, 1921 
/ 2 . 4 
l ’.7 
■ Average, 1.4. 2 Average, 0.55. 
These are the first analyses for silica for sea water off the North American 
coast. Unfortunately, the samples of water were not large enough to allow dupli- 
cate determinations except in two instances. 85 As the diagram (fig. 134) illustrates, 
the seasonal fluctuations proved much wider than Raben’s work would have sug- 
gested, with a pronounced maximum early in March, perhaps a second maximum in 
June and July, and something like six or seven times as much silica per liter at the 
beginning of March as in May or in autumn. That is to say, in the particular year 
in question (1921) the sea water near Gloucester was richest in silica a week or two 
prior to the time when we have usually found diatoms commencing to flower 
actively, became rapidly impoverished during the month when we have found 
diatoms most plentiful there in other springs, and poorest in silica at about the time 
the rich diatom flowerings come to a close. During June the supply of silica accu- 
mulated somewhat, and correspondingly we have twice found diatoms flowering in 
the bay late in summer or early in autumn (September in the year 1915, August in 
1922; pp. 394 and 391). With the seasonal fluctuations so notable for diatoms and 
fairly demonstrated for the concentration of silica, with the maxima for the former 
81 Doctor Wells writes me that although the iron of several of the containers was somewhat rusted, in most cases caroful analysis 
of the sediment showed practically no silica; and by analysis the iron of the cans was found to contain not more than 0.0002 gram 
silica per gram, so that measurable contamination of such large volumes of water by that agency is ruled out of consideration. 
