A strain of foot-and-mouth disease virus was recovered from a cow at the height of the disease, and propagated through at least 261 passages in guinea pigs. Considerably over 2000 animals proved susceptible, and the virus could be transferred at will back to cattle and hogs, and then again returned to guinea pigs. No natural immunity was discovered in guinea pigs. Secondary lesions were easily and regularly induced, thus making this strain particularly favorable for experimental purposes. In general, the guinea pig may be regarded as the animal of choice for laboratory studies. Guinea pigs could be infected by different methods of injection in different sites, but constant and regular production of primary and secondary lesions followed intradermal "tunneling" combined with subcutaneous inoculation of the posterior hairless pads of full grown animals. The virus was peculiarly epitheliotropic, giving support to the opinion that its portal of entry may be limited. The active agent purified itself of chance concomitant bacteria in the 1st passages, in a susceptible animal[long dash]a character possessed by filter-passing viruses in general. It was active in dilutions of 1:10,000,000, showing not only the minuteness of the active agent, but also the necessity for a change of technique from that employed with infectious agents of larger size. The dilution factor is also important in interpreting mere preservation of virus rather than multiplication. Furthermore, some samples of virus were not so active[long dash]a factor of 25 existed between the weakest and strongest samples among 15 titrated, thus indicating that comparative tests should be made with the same specimen. The incitant is not sedimented by centrifugation; failure of deposition is probably related to its minute size. The method of centrifugation has failed to remove "virucidal bodies" in the meaning of Frosch and Dah-men.