EXPERIMENT STATION BULLETINS. 635 



In the tests, two cubic centimeters of antigen were placed in each of four small 

 test tubes and the following quantities of serum added: 0.05, 0.025, 0.01, 

 0.005 c. c. The tubes were incubated either at 56° C for two hours in the 

 water-bath or in a dry incubator at 37° C for twelve hours. The final results 

 were practically the same for both methods of incubation. The results were 

 recorded as positive in complete clumping of the antigen, partial when the 

 supernatent Uquid was not entirely clear and negative in the absence of 

 clumping. 



HISTORY OF STRAINS 



Each culture of Bad. abortus employed in these studies except Strain 1, 

 has on some pre\aous occasion, produced characteristic lesions and enlarged 

 spleen in inoculated guinea pigs. The date of isolation and origin of Strain 1 

 is not knowTi, nor is there a past record which shows that it was at one time 

 virulent. The two strains. No. 1 swine and 2010, isolated from fetuses of abort- 

 ing swine are indistinguishable from Bad. ahoiius, Bang. The dates of the 

 isolation of the English and German strains are not known. The former 

 Enghsh strain was obtained from the Royal Veterinarj- College in England 

 and the German strain from Dr. Oldt of Berlin. When the two strains were 

 first brought to this country they produced pronounced lesions in guinea pigs. 

 Bad. melitensis 1 has no history of ever being virulent. Bad. melitensis 22 

 was classed as non-virulent. Bad. melitensis 27 as virulent by Dr. K. F. 

 Meyer of the University of California. 



RECORDS AND RESULTS 



As shown in Table II, the first comparative study was conducted in 

 January, 1920, using twelve strains of Bad. abortus which were isolated from 

 aborted fetuses or naturally infected milk from cattle. The second study 

 was conducted in April, 1921, and the third in August, 1921. There were 

 altogether twenty-one strains included in the second and third studies, twelve 

 strains of which were employed in the first stud5^ The three strains of Bad. 

 melitensis were introduced into the study for the purpose of comparing their 

 lesion-producing properties with those of Bad. abortus. 



In Table II, it is clearly showTi that of the first twelve strains employed; 

 five (1, 2, 3, 805 and 16) did not produce lesions in the first inoculation series, 

 nine (1, 2, 3, 805, J, E, C, Germany and England) did not in the second 

 series; and ten (1, 2, 3, 805, 7, 16, J, E, Germany and England) did not in 

 the third series. Strain 16 appeared non-\drulent in the first series, \'irulent 

 in the second and non- virulent again in the third series of inoculations. 

 Strain C, appeared virulent in the first series, non-\arulent in the second 

 and virulent again in the third series. There were four strains, as indicated 

 above, which proved to be non-virulent in all three series of inoculations. 

 Of the nine strains added in the second series of inoculations, two (Strains 

 No. 1 Swine and W) appeared to be non-virulent in the second series of 

 inoculations, and four of the nine (Strains No. 1 Swine, W, 2010, and 100) 

 appeared non-virulent in the third series of inoculations. There were two of 

 the nine strains, as indicated above, which proved to be non-^^rulent in both 

 series of inoculations. There were altogether five strains which failed to 

 produce lesions in guinea pigs throughout the studies. The strains were 1, 2, 

 3, 805, No. 1 Swine and W. Only one strain, 6, produced typical lesions and 

 enlarged spleens in all three successive series of inoculations, and only seven 



