ment for bacterial gill disease (Fish 1947) was developed at the Leavenworth (Washington) Station of the Fish and Wildlife Service in 1942 and was quite successful. Since then, Roccal has been used extensively; but because of variability in composition, its efficacy is not consistent (Rucker et al. 1949). The objection to the variability of Roccal was overcome by using another compound, pyridylmercuric acetate, which was suggested by Van Horn and Katz (1946) as having some therapeutic therapy. In 19.50 the fis'h toxicities of several of these compounds were determined for chinook salmon fingerlings (Oncorhynchus tshaw_ytscha) at the Entiat (Washington) Station of the Fish and Wildlife Service; and in 1953 the bactericidal activity was determined at the State Fisheries Hatchery, Issaquah, Washington, according to the methods described by Rucker (1948). The fish.were 2-inch, 3month-old chinook salmon fingerlings in an early stage of bacterial gill disease. The water temperature was 560 F. Here Lignasan at a concentration of 1:500,000 controlled bacterial gill disease experimentally. Because of the finding in this work, Lignasan was used, although to a limited extent, in a number of other State hatcheries in 1954. Lignasan was found to value. Pyridylmercuric acetate was tested be nontoxic, at the recommended concertexperimentally at the Leavenworth Station and was found to be very effective for bacterial gill disease. This compound had highly differential toxicities for bacteria and fish but was quite expensive (Rucker 1948, Burrows and Palmer 1949, Snieszko 1949). Another objection to pyridylmercuric acetate was its toxicity to rainbow trout--not to other species of trout or to salmon--at the concentration necessary to control the bacteria (Seaman 1950, Rodgers et al. 1951, Bryant 1951, Foster and Olson 1951).