(1)Statistical evidence cannot be deduced from such a small series of cases, but it would seem that the extent to which agglutinins for B. proteus X. 19 (Weil-Felix) were formed in the blood varied directly as the severity of the illness, since with the older and more ill patients a much higher dilution of serum agglutinated the organism at or near crisis than was the case with the younger patients, who were less acutely ill. An exception to this rule was met with in one case (Mrs E. R.), whose serum did not agglutinate the organism in a dilution beyond 1: 250, although she was the most severely ill of all the patients. It is possible that failure of agglutinin-formation resembles the absence of leucocytosis in the severely ill case of pneumonia and may be regarded as a bad prognostic feature.

(2)A positive Weil-Felix reaction was not obtained before the fifth day of illness, nor in any case before the characteristic rash appeared, so that the reaction is of little value as a factor in the early diagnosis of typhus. Later, a reaction in which a dilution of the serum of 1: 50 agglutinates the organism may be regarded as diagnostic of typhus.

(3)The Weil-Felix reaction remained positive for the longest period in those cases whose serum agglutinated in highest dilution, and in these patients a positive reaction (1: 25–1: 50) was still present nine to twelve weeks after infection, whereas in the milder cases, and in Mrs E. R., the reaction became negative in six to eight weeks from the onset of illness.

(4)In view of this heterologous reaction in typhus, it is interesting to record that the Wassermann reaction obtained before crisis was positive in four out of six cases. It was still positive in two of the most severely ill patients a week after crisis, but was negative in all cases a week later.