84 AMOEBAE LIVING IN MAN 



Encystation occurs in the same manner as in E. histolytica. In 

 typical cases, the precystic amoeba becomes rounded and secretes its 

 cyst wall, which also consists of a single layer, but is slightly thicker 

 than that of E. histolytica. Inside the cyst the nucleus increases in size, 

 and then divides into two. Each daughter nucleus again divides, 

 forming four nuclei ; and by a further division of each of these, 8 nuclei 

 are finally formed. A glycogen vacuole appears in the cytoplasm at an 

 early stage, and reaches its maximum development in the binucleate 

 cyst. It is then absorbed, and the mature 8-nucleate cyst is typically 

 without one. Chromatoid bodies may or may not be present at any 

 stage in development. All these stages are shown in PI. IV, figs. 56-62. 

 After this brief summary of the development of the cyst, I may now 

 consider several points in more detail, with special reference to the 

 differences between E. coli and E. histolytica — since these are of great 

 diagnostic importance. 



Size of Cysts. — Cysts of E. coli may be found with all diameters from 10 /a 

 up to 30 /i or even more. Cysts smaller than 10 /i have been described, 

 but I have never seen them, and it is not unlikely that these very 

 small cysts really belonged to a different species. They have all been 

 described, at all events, by observers unacquainted with the smaller 

 species living in man. Very large cysts — over 30 /x — are very uncommon, 

 and are generally supernucleate abnormal forms. The largest I have 

 ever found measured 33*5 fx, but Wenyon and O'Connor (1917) have 

 recorded a still larger specimen, measuring 38 /a X 34 /a. 



There can now be little doubt that E. coli, like E. histolytica, is a 

 composite species consisting of a number of different races distinguish- 

 able by the sizes of the cysts which they produce. This was suggested 

 by Wenyon and O'Connor (1917), and the measurements which I have 

 made leave no doubts in my mind as to the correctness of their view. 

 The question has recently been carefully studied by Matthews (1919), 

 who has reached the conclusion that there are at least three, and pro- 

 bably four, distinct races of E. coli, with living cysts of the following 

 mean sizes : 15 /a, i8'7 /j,, 21*7 /a, and probably 16*5 /a. In all these races 

 there is, of course, the usual variation in size around the modal value. 



Malins Smith (191 8) measured 1,000 cysts of E. coli from many 

 different infections, and found their average size was 17*3 fi. This 

 method, however, as I have already pointed out in considering E. histo- 

 lytica, throws no light upon the existence of races. From Smith's 

 figures, as Matthews (1919) points out, it seems probable that he studied 

 at least several distinct races. That these races actually exist in E. coli, 

 I think Matthews's figures show conclusively. He adopted similar 

 methods to those which I used for E. histolytica, and his results are 

 closely comparable. I believe, moreover, that the races of E. coli which 

 he has demonstrated by no means exhaust the number actually occurring 

 in this species. 



The measurements of E. coli cysts published by Kuenen and Swellen- 

 grebel (191 3) and Mathis and Mercier (1917 b) are too few to have any 

 value in this connexion. Their curves — based in each case on measure- 

 ments of only 100 cysts — are peculiar, but need not be further con- 

 sidered here. Their shortcomings have already been emphasized by 

 Smith (1918) and Matthews (1919). 



It has been proved beyond question that size is no certain criterion for 

 distinguishing between the cysts of E. coli and E. histolytica. The sizes 



