316 
BULLETIN OF THE BUREAU OF FISHERIES 
found both at the surface and at the bottom and probably were present at interme- 
diate depths. The larger number was sometimes at the surface and sometimes at 
the bottom. 
It will be seen from Figure 16 that the highest counts for the spring cruises of 1916 
were found in April and those for 1920 in March, but it must be noted that no samples 
were collected during April, 1920. Possibly, then, the maximum occurred in April, 
although we have no data to show it. 
The records indicate clearly, so far as the surface and bottom samples are con- 
cerned, that the high diatom counts, especially those of the spring cruises, occurred 
generally at the time of low salinity, but it does not follow that the low salinity caused 
the large increase in diatoms. 
DIATOM COUNTS AND HOMOHALINITY 
An interesting relation is that between the vertical distribution of salinity and 
diatom counts. In many cases during 1920 when high diatom counts occurred there 
were unusually close approaches toward homohalinity from surface to bottom. Such 
a condition has been observed by Nathansohn (1909) for certain regions in the sea, by 
Gran (1912) for the waters southwest of Ireland, and by others more recently. This 
relation existed quite frequently during the spring cruise of March, 1920, but in 1916 
this same condition was not so marked; in fact, usually the highest diatom counts 
were obtained on the April cruise while the nearest approach to homohalinity from 
the surface to the bottom, so far as we have records, occurred generally during the 
March cruise. 
DIATOM COUNTS AT RIVER MOUTHS 
The data which the survey has collected on the number of diatoms as a whole, 
not on separate species, do not supply any very convincing evidence for the theory 
that diatoms occur in greater abundance in the neighborhood of the mouths of rivers 
than at other places. Owing to the fact that quite a number of the diatom counts 
for the different areas along the western side of the bay were not made, a compara- 
tive study of the counts with reference to the river mouths on that side does not 
yield very satisfactory results. However, it may be seen from the surface data in 
Table 10 that, so far as our data go, the numbers were usually comparatively large 
at one or the other of the areas P, J, or H. These areas in the case of P and J are 
close to the mouth of the Potomac River and area H is near the entrance to the Rap- 
pahannock River. On the other hand, during the March cruise while the counts 
were high at area P near the mouth of the Potomac River they were even higher 
farther up the bay. 
Table 10. — Diatom counts in surface samples on west side of bay, per liter 
1920 
1921 
January 
March 
May 
July 
August 
October 
December 
January 
14. 700 
7,400 
243, 100 
11, 100 
( 3 ) 
4, 600 
(') 
86, 300 
(>) 
465, 400 
690, 100 
769, 200 
( 3 ) 
269, 100 
593, 700 
507, 300 
(») 
( 3 ) 
7, 400 
7, 800 
3, 400 
10, 5, 800 
14,100 
9,900 
900 
3,700 
3. 900 
4. 900 
( 3 ) 
1, 000 
5,700 
3, 200 
10,600 
( 3 ) 
3, 900 
( 3 ) 
28. 300 
31.300 
13. 400 
5, 700 
(») 
(•) 
( 3 ) 
36,300 
29,000 
11,800 
0) - 
12,200 
i No sample. 
> No count. 
