B A CTERIOL OGICA L STUDY OF WA TER. 639 



apparatus may be employed for counting the colonies 

 in rolled tubes. It consists in dividing the tube by 

 lines into four or six longitudinal areas, which are sub- 

 divided by transverse lines about 1 or 2 cm. apart. The 

 lines may be drawn with pen and ink. They need not 

 be exactly the same distance apart nor exactly straight. 

 Beginning with one of these squares at one end of the 

 tube, which may be marked with a cross, the tube is 

 twisted with the fingers, always in one direction, and 

 the exact number of colonies in each square as it 

 appears in rotation is counted, care being taken not to 

 count a square more than once ; the sums are then added 

 together, and the result gives the number of colonies in 

 the tube. This method may be facilitated by the use 

 of a hand-lens. 



In all these methods there is one error difficult to 

 eliminate : it is assumed that each colony has grown 

 from a single organism. This is probably not always 

 the case, as there may exist clumps of bacteria which 

 represent hundreds or even thousands of individuals, 

 but which still give rise to but a single colony ob- 

 viously this is of necessity estimated as a single organ- 

 ism in the water under analysis. 



Where grounds exist for suspecting the presence of 

 these clumps they may in part be broken up by shaking 

 the original water with sterilized sand. 



What has been said for the bacteriological examina- 

 tion of water holds good for all fluids which are to be 

 subjected to this form of analysis. 



THE SEWAGE STREPTOCOCCUS. Houston 1 reached 

 the conclusion that there is constantly present in sewage 

 a particular form of streptococcus which is really more 



1 Houston : Ann. Report, Local Gov. Board, xxviii. 



