5© Marshall O. Leighton 



purposes of argument, that they may be exceptional, the contentions 

 of the writer are not damaged thereby. It should be remembered 

 that in making sanitary analyses we are not developing a scientific 

 theory that must stand or fall according to the weight of cumulative 

 evidence for or against, but we are trying to determine whether 

 or not the use of that water will cause sickness and death. This is 

 a positive purpose ; and if it is admitted that there can be exceptions, 

 even though they be few, the whole scheme of analytical procedure 

 fails of that purpose. Exceptions are not' predestined, and in this 

 case cannot be guided or defined. The chemist who calls a water 

 ''good" upon the evidence presented by his nitrogen determinations 

 has no means of knowing whether or not this water may be one of 

 the exceptions. Supposing it be polluted like the Lake Champlain 

 samples in Series "B," and a family or a community accepts the 

 favorable opinion of the chemist and is stricken with typhoid fever 

 — think you that that chemist will be justified by appearing before 

 those bereaved relatives and reciting the fact that the great mass 

 of evidence sustains certain bases of interpretation, and the scat- 

 tering exceptions do not, from a scientific standpoint, destroy the 

 integrity of the theory ? The most befitting remark at this junc- 

 ture seems to be the old proverb: "A chain is no stronger than its 

 weakest link." 



After all, perhaps the strongest indictment of the sanitary analy- 

 sis is that it is unnecessary for the purposes for which it is generally 

 used. No one will question its value in sewage experiments, but 

 when the purpose is to determine whether or not a water be potable, 

 there are more satisfactory ways of solving the problem than by 

 making the conventional grind of nitrogen determinations, even 

 though it be admitted for the moment that those determinations 

 fulfil all the great purposes claimed for them. 



It may be accepted as axiomatic that no river, upon the drainage 

 area of which there is any population, will furnish a water fit for domes- 

 tic consumption in its raw state. That this shall hold true it is not 

 necessary that the population shall be gathered into cities and be 

 provided with sewerage systems. Rural population, even though 

 widely scattered, is dangerous. Wherever people live along the banks 

 of a stream there will always be dangerous pollution. Indeed, the 



