SIO Reference 67-11. Biologists who are interested in the primary productivity of the ocean are often unable to make productivity determinations by the "in situ" technique because of the demands of other groups on the ship's time. Under these conditions it has been the practice to use a substitute technique called the "deck incubator" technique, in which the sample of plankton-bearing ocean water is transferred to a deck incubator and allowed to propagate for an appropriate length of time before determining the 14C uptake.From a radiometric point of view these two techniques differ to a considerable degree. In the "in situ" technique the plankton-bearing ocean water is returned to its original depth location where propagation takes place at normal radiant energy levels and with normal spectral and geometric distribution of the radiant energy. In the "deck incubator" technique the radiant energy level available is usually much too high and must be controlled, the spectral distribution and band width are abnormally different and the geometrical distribution of the radiant energy has no similarity whatsoever with the natural geometrical distribution underwater.Experimental results for any individual case are still further confused by the fact that the measured radiant energy must be sensed by a detector, usually of the photovoltaic type, having spectral and directional sensitivity properties unrelated in any way to those of the phytoplankton which are being studied, and magnitude response which is often arbitrarily nonlinear.These problems have been recognized to some extent by various workers engaged in primary productivity work and effective measures have generally been taken to control the magnitude of the radiant energy in deck incubator determinations. However, not enough has been done to control its spectral distribution, and very little, if anything, has been done to control the geometrical distribution of the radiant energy or to design a suitable photodetector or to design an appropriate deck incubator.Perhaps the most curious effort, in connection with the control of the spectral sensitivity of the photocell-filter combination, has been the deliberate selection of photodetectors which measure "lumens", and the general adoption of "lumens" as a radiometric unit for productivity work. The "Lumen" is, of course, a unit which was specifically orginated for radiant energy measurements relating only to human vision at the levels of photopic response. Its use in connection with photosynthesis is meaningless and is so recognized by many biologists who nevertheless continue to measure and publish in "lumen" units.From the point of view of international cooperation, the communication of results is exceedingly difficult. Not only do we have measurements of radiant energy which are meaningless for correlation with productivity, but every laboratory seems to have a somewhat different type of meaningless measurement.