1887.] Prof. C. Ewart on Bacteria in Lymph, &c., of Fish. 263 
to confirm to a certain extent Olivier and Richet’s observations. 
Although I had often examined microscopically the blood and 
tissues of fish, it was not until recently, when at work in the Oxford 
Physiological Laboratory, that I was convinced that bacteria are 
often present in immense numbers in the peritoneal fluid, and in 
smaller numbers in the blood of apparently healthy fish. 
I first noticed bacteria in the blood of a roach ( Leuciscus rutilus). 
This roach, for some hours before it was taken from the water, had 
been occasionally swimming on its side at the surface, — an indica- 
tion that it was in an exhausted condition. Immediately after 
the fish was killed, a drop of blood taken from the heart by a 
sterilised pipette (with all the necessary precautions) was found to 
contain a considerable number of slender motionless bacilli 
measuring from ’003 to *008 mm. in length. On an average four 
bacilli were visible in the field at a time with Zeiss’s P objective 
and No. 1 eyepiece. The peritoneal fluid, which was next examined, 
contained so many bacilli that it was impossible to count them ; 
the bacilli were usually lying amongst large granular lymph cells, 
and they were longer and more slender than those in the blood. 
Similar bacilli were found in the lymphatics, spleen, liver, and 
kidney, and they were abundant in the muscles in contact with 
the peritoneum ; while very few were found in the muscles under 
the skin of the trunk, and still fewer in the muscles of the tail. 
The intestine was crowded with similar bacilli to those found in the 
body-cavity, and in addition there were a number of large and small 
bacteria and micrococci. Bacilli were also found in the walls of the 
intestine and in the bile duct. 
Believing that there was some relation between the diminished 
vitality of the above roach and the numerous bacilli in the tissues, 
I examined a considerable number of healthy roach and also other 
fresh-water fish, e.g ., trout ( Salmo levenensis), perch ( Perea Jiuvia- 
tilis ), carp ( Cyprinus auratus), and eels ( Anguilla vulgaris). In all 
the healthy specimens examined, with the exception of the trout, 
bacilli were found in the body-cavity. Bacilli were also present in 
the blood of the carp, and on one occasion four bacilli were detected 
in a drop of blood from what appeared to be a healthy roach. In 
some the peritoneal fluid contained numerous bacilli, while in others 
only a few w T ere visible • generally there was some relation between 
