The rates at which Calanus finmarchicus eat both carmine particles and the diatom Nitzschia closterium have been investigated by Fuller & Clarke (1936) and Fuller (1937). They found that the number of Nitzschia or of carmine particles, eaten by a Calanus in unit time, was proportional to the concentration of Nitzschia or of particles in the water. Lucas (1936) also found that Neomysis and Eurytemora ate Nitzschia at rates which were roughly proportional to the concentration of diatoms in the water over fairly wide limits.

During 1935 I had made, as part of another investigation, three experiments on the rate at which Calanus eat diatoms of larger size. This happened much more rapidly than Fuller found to be the case with suspensions of Nitzschia. Meanwhile Lowndes (1935) had concluded, from direct observations, from the anatomy of their mouth-parts and from observations by Lebour and Marshall, that Calanus should be able to catch and select food. He concludedthat Calanus does not depend entirely upon its (automatic) filtering mechanism for all the food it obtained.

The quantitative data which I had collected in these three experiments, in conjunction with those published by Fuller & Clarke (1936,1937), were suitable for examining this contention. They suggested that the animal when fed with diatoms of moderate size selected by catching the species it preferred, while it automatically filtered the minute Nitzschia. With these possibilities in view, experiment N 90 was made in order to link up the three experiments with the numerous experiments made by Fuller (1937).