282 CLINICAL BACTERIOLOGY AND H^LMATOLOGY 



taken into account; a complication (e.g., perforation) which 

 causes a rapid and marked leucocytosis in a patient who has 

 not been greatly enfeebled by a long and severe illness may 

 cause no increase, or even a diminution, in an exhausted sub 

 ject. If this rule is forgotten, the indications from the blood- 

 count may be unjustly stigmatized as misleading. 



Perforation. There is a rapid increase of leucocytes, which 

 is said to occur in half an hour; the total number may be 

 15,000 or more. In a patient who is not greatly exhausted 

 this is a very sure sign; exceptions do occur, but are not 

 frequent. 



Hemorrhage. If a count has been made a short time pre- 

 viously, a fall in the red corpuscles may be noticed in some 

 cases, but does not seem always to occur. There may also 

 be slight leucocytosis; this is only of importance in that it 

 leaves the diagnosis of perforation or haemorrhage doubtful, 

 though raising the presumption that one or other has 

 occurred. As a rule it is not high, and the more it rises above 

 15,000, the more likely is perforation to be the cause. 



Pneumonia. When it occurs early in the disease there is 

 usually a slight leucocytosis; when it occurs later there may 

 be none. Do not exclude pneumonia, therefore, because of 

 its absence. 



It follows from the above that a raised leucocyte count in 

 typhoid fever always points to a complication of some sort, 

 but does not necessarily indicate its nature. 



PNEUMONIA. Here the results of blood examinations are 

 fairly constant, and of much value in diagnosis and prognosis. 

 In ordinary cases there is marked leucocytosis, due entirely to 

 an increase of the polynuclears, which may reach 95 per cent. 

 There is also moderate secondary anaemia. 



According to Ewing, who has had much experience on the 

 subject, the grade of leucocytosis is roughly proportionate to 

 the extent of the lesion. The following are his averages : 



Average Leucocytes. 



1 lobe affected ,*/ 22,000 



2 lobes 22,700 



3 2 5,ooo 



4 2 7>oo 



But higher counts, often much higher, are frequently seen. 

 These figures serve to exclude typhoid fever, tuberculosis, 



