EXPERIMENT STATION BULLETINS. 653 



were selected for study because they were kuown to be badly infected. 

 Another thing worthy of notice in this table is that the infection is 

 either present in a high percentage of the animals of a herd or is not 

 present at all, which condition may be due to the very infectious nature 

 of this disease of cattle. 



To show the duration of Bad. abortus infection when once established 

 in the cow's udder tests have been made at intervals of six weeks cover- 

 ing a period of three years. These are tabulated in Graph I. 



It will be noticed that once the infection becomes established the 

 milk remains a possible source of infection for years. In no cow under 

 observation has the infection died out after becoming firmly established. 



In hopes of throwing some light upon the manner in which the udder 

 l)ecomes infected with Bad. abortus the results shown in Graph II were 

 tabulated to show the first appearance of the antibodies in the milk of 

 the various quarters. In some instances it will be noticed that the re- 

 action is first shown to appear in two quarters at the same time. This 

 is probably due to the fact that as the milk was tested only at intervals 

 of six weeks tliore was time for the reaction to appear and spread from 

 quarter to quarter before it was again tested. In several instances 

 v.iiicli are not shown in this Graph the reaction had spread to all four 

 {[uarters when tested after a regular negative test six weeks previously. 

 In these instances we have no way of telling which quarter was the first 

 to show the reaction. 



In Graph II it will be noticed that one or the other," or both of the 

 rear quarters were tlie first to show the reaction in five of the eight 

 cases. In the three other instances, one of the rear quarters and one of 

 the front quarters show the reaction at the same test. As the front 

 quarters of two of these cows gave a very weak reaction while the rear 

 quarters were very strong it seems safe to conclude that in these two 

 instances the rear quarters were first to show the infection also. In the 

 other instances it is impossible to say definitely whether the left rear 

 or left front quarter was the first to show the reaction. 



The fact that the rear quarters are the first to show this infection 

 points to the genital discharges as the probable source of the infection. 

 In infection of the genital tract the discharges are often found to con- 

 tain Bad. abortus in a virulent form. Through switching the tail which 

 often becomes coated with this infectious material, or because of the 

 discharges running down the back of the udder the rear quarters would 

 naturally be the first to become infected. The infection could then be 

 spread to the other quarters upon the hands, during the milking process. 

 These facts seem to point to the genital discharges as the original source 

 of the infected udder. 



To prevent this infection from spreading to the udder and thus making 

 the milk unsafe by becoming a possible disseminator of the infection, 

 steps should be taken to prevent the discharges from coming in contact 

 with the teats. 



The age of a number of the cows that developed Bad. abortus infected 

 udders while under observation is tabulated in Table II. 



