In situ absorption data collected with a reflective-tube absorption meter are presented. Various procedures for correcting the data for scattering error and for extracting chlorophyll absorption from the raw signal are explored. Based on our knowledge of the distribution of particle types and on measured backscattering, the scattering correction as a function of total scattering is found to vary significantly with depth. However, absorption in the near infrared is shown to be highly correlated with backscattering. We thus postulate that this signal is mainly due to the scattering error and possibly to absorption by dissolved substances and particles of a detrital nature. Thus, the infrared signal seems to provide a good correction for the measured a(676) to obtain chloropigment absorption. Indeed, a(676) - a(750) (corrected for water absorption and 750 temperature dependence) was found to correlate strongly with fluorescence. However, a(750) was found to be highly temperature dependent, so a(712) was chosen for this purpose in future measurements. An examination of optical microstructure in East Sound, Orcas Island, Washington showed numerous peaks with vertical dimensions of the order of tens of centimeters. These peaks may contain the majority of the biomass in the system. The relative magnitude of the a(676) - a(712) and the a(712) signals varied greatly from one peak to another and systematically with depth, presumably reflecting the nature and physiological states of the populations in the various peaks.© (1992) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.