PLATES XIV AND XV

SWARMING by proteus bacilli, first described by Hauser (1885), has continued to puzzle microbiologists. It is thought to be caused by negative chemotaxic metabolites (Lominski and Lendrum, 1947; Hughes, 1957; Hoeniger, 1964; Brogan, Nettleton and Reid, 1971) that accumulate during growth. When a threshold concentration is reached, division is inhibited and the bacteria develop into highly motile long forms which swarm down the concentration gradient. On escape from the inhibitory concentration, the long forms fragment into short rods which multiply until the metabolites again accumulate to the critical swarming concentration. The process is then repeated and results in the well-known pattern of concentric circles seen on certain media. These chemotaxic metabolites have not been isolated and their chemical composition is unknown (Hughes); they appear to be volatile (Lominski and Lendrum; Hughes) and to be inactivated by ether (Lominski and Lendrum). Hughes showed that proteus bacilli also form stable non-volatile substances which may be responsible for the line of demarcation (Dienes, 1946, 1947) between swarms of different strain