78 
BULLETIN OF THE BUREAU OP FISHERIES 
In Figure 18 all of these data are presented graphically. There are two main 
series, the samples from Offatts Bayou and those from West Bay. In the former 
each point represents the average of three samples taken at the three places designated 
in Figure 5 (stations 1, 2, and 3). This average is a fair estimate of the abundance of 
diatoms in the bayou, for the places where collections were made include not only the 
water near the mouth which is changed frequently by tides but also that farther up 
in the bayou where the water is only slowly exchanged. In the case of the points 
representing West Bay collections each point is the average of three determinations; 
two from near the Deer Islands and one from a short distance out in West Bay from 
the mouth of the bayou. (Fig. 4, stations 4, 5, and 6.) These points do not give as 
complete an idea of the abundance of diatoms in West Bay as do those referring to 
the bayou, for West Bay is much larger. In spite of this fact, however, it will be 
observed (fig. 18) that there is a striking parallel between the records in the two 
places. 
Another series of samples is also represented in the figure, referring to East Bay 
near Hanna Keef. Samples were taken only infrequently, and many of those taken 
are not included because the water was muddy and clogged the net, making the sam- 
ples of doubtful value. However, these few points are distinctly in agreement with 
those of the other two series, suggesting that the records constitute a fair estimate of 
the diatom content of Galveston Bay as a whole. 
The most striking characteristic of these records (fig. 18), when minor fluctua- 
tions be overlooked, is the gradual reduction in abundance of diatoms in the water 
from the end of March, when the first samples were taken, to the end of August, when 
sampling was discontinued. In the former case the water was so full of diatoms that 
a haul had to be shorter than the standard time, due to clogging of the net. Toward 
the end of July and during August most of the samples contained little else save 
oyster larvae. In most cases at this time very careful microscopic examination was 
necessary in order to find one or two diatoms. These samples were designated as 
halfway between none and very few and actually represent the virtual absence of 
diatoms of the sizes taken by the net. 
Comparison of the graph of diatom abundance with those of temperature (fig. 9) 
and salinity (fig. 10) is of interest. The temperature rose rapidly toward the end of 
March to over 25° C. and after this time diatoms became rapidly less abundant. 
The diatom records for the end of April and beginning of May are too scattered to be 
particularly significant, but the few samples during this period showed extremely few 
diatoms in the water. The salinity during March, April, and half of May was in the 
neighborhood of 20 parts per thousand. However, following May 20 the salinity 
dropped quickly from 20 to about 14 parts per thousand, due to increase in river 
discharge. This was accompanied by a drop of a few degrees in temperature. At 
the same time there was a great increase in the diatom content of the surface water 
which lasted for about 10 days, and then abruptly the diatoms disappeared. 
The variations in the average temperature of bayou water (fig. 9) after the middle 
of May appear to bear a definite relationship to the abundance of diatoms in the 
samples taken in this place. In nearly every case, between this time and late in the 
summer, whenever there was a marked fall in water temperature there was a rise in 
abundance of diatoms collected. When the temperature rose again the diatoms taken 
became less abundant. Even minor variations in temperature were accompanied 
by fluctuations in the abundance of diatoms taken. Whether the temperature 
variation is itself responsible for differences in the abundance of diatoms is not clear. 
