40 



Matthews and Smith : but it must be remembered that the findings 

 are based throughout on the results of only one examination per 

 case. None of the 2,137 persons in the whole series had ever been 

 abroad. 



If we now take the results of the newer Reports, given on 

 earlier pages, and express them in the same way — converting 

 percentages into actual numbers as before, when necessary — we 

 obtain the figures shown in the following table : 



Protozoa. 



If we add these findings from 1,009 individuals who have 

 never been abroad to those from the 2,137 similar cases in the 

 Liverpool series, we obtain the following results : 



Protozoa. 



These final percentages, based on the records obtained from 

 the examination of over three thousand individuals — men, women, 

 and children — who have lived all their lives in Britain, give us 

 material for making an approximately correct estimate of the 

 extent to which the indigenous population of (Jiese islands is 

 infected with intestinal protozoa. But they cannot be accepted, 

 as they stand, as correct ; and for the following reasons. 



The greater part of the series (over 2,000 persons) is composed 

 of the Liverpool cases. These were all examined but once apiece, 

 and they thus form — in this respect — a homogeneous series. The 

 other series, however (over 1,000 oases, or roughly, one-third of 

 the whole), consist of persons examined a variable number of 

 times each, and consequently the whole series is not homogeneous 

 in respect to the number of examinations made per case. It is 

 well known — from the figures already collected and published by 

 myself (1917) and others — that the number of infections found 

 in any series examined depends upon several factors : and one of 

 the most important of these is the number of examinations made 

 per case in the series. (I leave out of account now the other 

 factors — such as the competence of the examiner, the time taken 

 in making the examination, &c. — since there is no reason to 



