1. Experiments have shown that tracheal exudate from two strains of laryngotracheitis of chickens from New Jersey and two from California when suspended in bouillon and passed through Berkefeld V filters will produce the disease. Two of six Berkefeld N filters allowed the etiological agent to pass, whereas four did not. Attempts to produce the disease with Seitz filtrates were unsuccessful. These results demonstrate that laryngotracheitis is caused by a filtrable virus that because of its size or some other property does not pass readily through the finer filters.
2. It has been shown that the sera from fowls that have recovered from an infection with one of the New Jersey viruses will neutralize the same strain and also the one California strain tested. In order to demonstrate neutralization conclusively it was necessary to titrate samples of dried virus and in the tests to use approximately ten infecting doses.
3. The virus dried over calcium chloride for 10 days and then stored in the refrigerator for 60 days produced disease. Kept over calcium chloride for a month it was still active and when dried by Swift's method it remained alive for 5 months.