!9 2 °] Allen: Plankton of the San Joaquin River 57 



Navicula bacillum Ehrbg. 



Station I Station II Station III Daily Hourly 



Average 56,910 169,399 42,869 16,079 65,103 



Identification doubtful. Percentage retained by net must be very 

 small. It may seem a waste of time to list so many diatoms of uncer- 

 tain identity, especially of the smaller forms, but it is the writer's 

 opinion that to do so may serve two purposes : first, to give a faint idea 

 of the large number of forms present ; second, to show how very many 

 forms there are beyond the reach of ordinary quantitative methods. 

 For example the small diatom listed under this present heading was 

 quite probably present in one hundred times the numbers recorded, 

 possibly more. It appears in April at all stations and is erratic in 

 appearance and numbers after that. The maximum appears in Sep- 

 tember at Station I and Station III but in June at Station II. In the 

 circumstances no safe conclusions can be drawn from such a small 

 organism. 



Navicula didyma Ehrbg. 



Identification doubtful. Losses through net heavy. Recorded once 

 at Station I, ten times at Stations II and III, always in small numbers. 



Navicula dubia Ehrbg. 



Identification doubtful. Losses through net heavy. Recorded six 

 times at Station I, four times at Station II and five times at Station 

 III in small numbers. 



Navicula gracilis Ehrbg. 



Station I Station II Station III Daily Hourly 



Average 209,568 722,906 1,514,448 166,204 1,177,953 



Identification uncertain. Losses through net very heavy. Re- 

 corded in every month at all stations. Maximum in August at all 

 stations on a rather abrupt rise in numbers. Certainly favored by 

 higher temperatures and quiet water but hindered by sewage. These 

 points are shown by the remarkable development at Station III with 

 quiet water but with little sewage and by the fact that the large num- 

 bers shown by Station II through May probably came from the 

 washing out of some sloughs where there had been a time of quiet 

 just preceding. Recurrent pulses are fairly well marked. This form 

 is evidently very important numerically. The count may include 

 several species. 



