Double infection of the allantoic sac with Melbourne and WSN viruses induced the formation of a combination virus, which had some of the antigenic properties of both parents and which maintained itself through serial passage in the chorioallantoic sac. In the course of prolonged passage in the egg, three varieties of combination virus were found. The first (X1) occurred in ten passages and was characterized as follows: X1 was produced in and could be successfully passed from those chick embryos that had received very large inocula of virus. The X1 hemagglutinin was efficiently inhibited by both M and W antisera. Each passage fluid, containing X1 as the predominant strain, also contained large amounts of the parent forms M and W. It is very likely that X1 reverted to parent types of virus at a high rate.
The second variety (X2) arose from the passage series of X1 and was carried for approximately 20 passages without definite alteration in its character. It differed from X1 in that it had the ability to appear as the predominant form occasionally in embryos that had received a limiting infective dilution of virus. Because of this, it seems probable that X2 reverted to parent types at a slower rate than X1. Like X1, this virus was never obtained in pure form and suspensions in which it predominated contained large amounts of M and W, especially the former. Good evidence was obtained that X2 reverted to M and W virus while multiplying in the chorioallantoic sac.
X3 was derived from passage of X2 virus after 32 transfers and can be characterized in several ways: (a) X3 yielded the only X3 fluids on passage in eggs. This was in striking contrast to X2 which could not be passed without giving rise to some fluids in which M and W predominated, (b) The X3 hemagglutinin was weakly inhibited by all but the highest concentrations of M serum, but was strongly inhibited by W serum. X2 was readily inhibited by relatively small amounts of both sera. (c) X3 virus, after several passages at limiting dilutions, was neutralized to a highly significant degree by specific M and W sera. This is not necessarily an essential difference from the behavior of X2 since technical difficulties (large amounts of M and W in X2 fluids) may have prevented the demonstration of equally striking double neutralization of this virus, (d) After five limiting dilution passages, the X3 virus retained the characteristics listed under (a) to (c). It is felt that the foregoing facts justify the conclusion that X3 is a stable virus which combines some of the specific antigens of two parent types.