In previous work in this laboratory, Mycoplasma suipneumoniae was recovered in liquid medium from 13 % of individual cases and 18 % of outbreaks of enzootic pneumonia in pigs. In the work now described, however, these recovery rates, when judged by the same criteria, were 45 and 75 %, respectively. As there was evidence to suggest that this second series of pneumonic cases was less suitable for cultural examination than the first series, some of the other factors that might have improved the recovery rate were investigated.

Some improvement was probably achieved by inoculating the liquid medium with three or four different dilutions of pneumonic tissue, each dilution always being in duplicate, and by incubating the inoculated tubes for over 3 weeks before discarding them.

A second advantage could have derived from the fact that all batches of liquid medium were tested for their ability to support the growth of M. suipneumoniae before being used to culture field material.

The effect of varying one constituent at a time was observed in controlled experiments: different batches of pig sera had a marked, variable effect on the growth of both M. suipneumoniae and Mycoplasma hyorhinis; medium made with purchased Hartley's broth was found to be superior to medium incorporating broth made in this laboratory, more so for the growth of M. suipneumoniae than M. hyorhinis; the incorporation of yeast extract made in this laboratory gave a marginal advantage for the growth of M. hyorhinis; and both mycoplasmas grew equally well in medium containing or lacking thallium acetate.

Some batches of medium were, by chance, markedly selective for the growth of M. suipneumoniae compared with M. hyorhinis. As the full reasons for this were not known, attempts were made to develop selective media in a more direct way. One such medium contained 5% pig serum and 15% horse serum, and a second was of similar composition, except that the pig serum used inhibited preferentially the growth of M. hyorhinis compared with M. suipneumoniae. Both media markedly favoured the growth of M. suipneumoniae when tested separately with cultures of M. suipneumoniae and M. hyorhinis. The second medium yielded M. suipneumoniae when inoculated with a 10-1 dilution of a culture of M. suipneumoniae and a 10-2 dilution of a culture of M. hyorhinis, whereas a standard batch, of liquid medium, similarly inoculated with M. suipneumoniae did not yield this mycoplasma until the M. hyorhinis culture included in the inoculum was diluted to 10-6.

Both selective media, when tested on a small number of field cases, gave improved isolations of M. suipneumoniae compared with the routine batches of liquid medium used initially.

Considerable difficulty was experienced in producing a sufficiently high level of antibodies to M. hyorhinis in pig sera and to M. suipneumoniae in rabbit sera. This exacerbated the problem of isolating and identifying M. suipneumoniae from field cases of enzootic pneumonia by this cultural method.