but has demonstrated its ability to catch most of the larger forms in spite of the comparatively small opening.
The remaining source of error, due to sampling, fractioning and counting, is probably not great and, with ordinary care and an occasional recheck, can be reduced to a minimum. Rather elaborate statistical studies of errors in plankton collection and enumeration have been undertaken by Ricker ( 1937a) and other Canadian limnologists. These studies, while they are of great interest, seem to be of less practical value to the limnologist. Sampling of a body of water is not assumed to be an exact science, and this has well been illustrated in some remarks by Allen ( 1938) in commenting on the accuracy of certain plankton catches in the Gulf of California. He states, "Under such considerations it should be apparent that the rigid accuracy more or less erroneously attributed to all scientific activity is actually unattainable and that the investigator must be content with something less than the popular ideal. As a matter of fact, this condition is not as bad as it seems or as it may be made to appear to be. The life of the investigator himself is not a result of rigid accuracy. It is a series of compromises, readjustments, and approximations. In fact, all life is like that to a greater or lesser degree and even inanimate Nature is not free from appearances of uncertainty or confusion. And, if a determination of the characteristics of a population in the ocean could be made with rigid accuracy for a given instant, the terms expressing this accuracy might be widely misleading if applied to the population present an hour or a day later. . . . To really know Nature one must sacrifice something of mathematical accuracy at times and depend upon a series of approximations (which, paradoxically, will lead to a nearer approach to accuracy than would be possible by the use of extreme refinements).
The examination of stomach contents has shown that many fish utilize plankton as an important element in their diet. It has generally been believed that fish simply strained a great quantity of water and separated out whatever was in it. Some recent work, however, has shown that in some fish, at least, this is not true and that there is an actual selection of food by the fish. Battle et al. ( 1936) found that the herring captures each individual plankter. This can also be noticed in goldfish when Daphnia are put in the aquarium. Every plankter is hunted down by individual capture. Ricker ( 1937) found that the sockeye salmon seemed able to distinguish between individual plankters and even showed individual preferences. He confined 10 fish in the same limited environment and found that seven of them took almost all Daphnia, three took Cyclops as well, while one of the three took only Cyclops! Ricker points out that individual preferences for different organisms may cast some doubt on the reliability of average food consumed in series containing a few stomachs only. The kind of food eaten was also found to vary with the season; in May and June, when the fish were small, Cyclops and Bosmina were fed upon. Later in the summer the food was mainly Daphnia, and in winter Cyclops and occasionally considerable numbers of Bosmina were taken. The sockeye salmon continued to feed mainly on plankton to the third year.
During the course of this discussion a number of lines of investigation, in which much more work can be done, have been pointed out. In summary these are as follows: further study of zooplankton organisms and their distribution; pioneer quantitative work on the plankton of little known regions, particularly in the tropics and in northern regions, more work on seasonal distribution of plankton and organic matter, especially in unknown regions; further study of seasonal changes in the diurnal migrations of zooplankton and the development of a "perfect" collecting device for macro-plankton.
ALLEN W. E. 1938. "The Templeton Crocker Expedition to the Gulf of California in 1935--The"Phytoplankton