250 BACTERIA PATHOGENIC TO MAN 



ever, and some along the small intestine, were hemorrhagic. In Long- 

 cope's case the intestine showed no changes either on gross or micro- 

 scopic examination. The spleen in both cases was enlarged. The 

 other pathological changes were those common to febrile conditions. In 

 Tuttle's case a few erosions just above the ileocsecal valve were present. 



SOURCE OF INFECTING BACILLI. Tuttle's case happened to be a 

 laboratory employe in the service of the Department of Health and 

 was carefully investigated by us. We found that two families consisting 

 of eleven members drank water from an open uncovered tank. During 

 the summer the tank was not cleaned and was only occasionally filled 

 by pumping in water from the city supply. Sometimes the water was 

 the color of tea. During a single week four members of one family 

 and three of the other were stricken with a typhoid-like fever. The 

 two families had no social intercourse with each other. 



Symptomatology. It is a significant fact that many of the reported 

 cases of paratyphoid infection were considered to be genuine typhoid 

 fever without the Gruber-Widal reaction until a bacteriological study 

 revealed their true nature. 



Intestinal hemorrhages, furunculosis, initial bronchitis, cystitis, 

 pyelonephritis (?), purulent arthritis, bronchopneumonia, and venous 

 thrombosis have been reported as complications. Osteomyelitis is the 

 only recorded sequel. 



The duration of the disease has varied from twelve to eighty-four 

 days, with a majority of the cases continuing between twenty and 

 thirty-six days. Some of the cases have been of short duration, lasting 

 from twelve to eighteen days. 



The Serum Reaction in Cases of Paratyphoid Infection. Since the 

 introduction of serum reactions as a means of diagnosis, it has been 

 a well-recognized fact that a small proportion of cases which are clinic- 

 ally typhoid fever fail to give the reaction. Brill, adding to Cabot's 

 statistics, finds that of 4879 cases 4781, or 97.9 per cent., gave the reac- 

 tion. Gwyn gives 99.6 per cent, as the percentage of positive reactions 

 in the Johns Hopkins Hospital. On the contrary, in most of the reported 

 cases of paratyphoid infection a reaction, except with low dilutions, 

 against the bacillus typhosus has been absent. It is probable, then, 

 that some at least of the typhoid cases with negative reaction were 

 really paratyphoid infection. 



On the other hand, it cannot be assumed that all cases clinically 

 typhoid fever, which have been reported as giving the Gruber-Widal 

 reaction, were cases of genuine typhoid infection. The brilliant work 

 of Dunham on the typhoid-colon group of bacilli and its serum reac- 

 tions has brought out the fact that certain members of this group may 

 be mutually interacted upon by sera of infected patients and of 

 immunized animals. This is especially true of sera in low dilution. 

 Since in the earlier years of the Gruber-W T idal reaction the technique 

 had not been worked out, and dilutions were more frequently low than 

 not, some of the cases reported as typhoid fever may have been infec- 

 tions with paratyphoid bacilli. 



