January 19, 1900.] 



SCIENCE. 



107 



each inhabitant, a consumption about three 

 times as great as that of American cities. This 

 statement is traced by Mr. Plerschel to certain 

 hypothetical computations published by Prony 

 in 1817, which are shown to rest upon unwar- 

 ranted assumptions. A probable value of 

 the quinaria is found in three different ways : 

 first, by computations from the measured cross- 

 section of an aqueduct now standing and the 

 ancient slope of its water surface as marked by 

 the incrustations of the deposited limestone ; 

 second, by actual gaugings of those ancient 

 aqvieducts that are now in use ; and third, by 

 computations from reasonable data of the dis- 

 charge of pipes which delivered water to houses. 

 An analysis of the work of Blumenstihl and 

 Belgrand leads to the conclusion that the value 

 of the quinaria was somewhere between 2500 

 and 9000 gallons per day. Accepting the state- 

 ment of Frontinusthat 14,000 quinarias were de- 

 livered within the city, and calling its popula- 

 tion one million, it follows that the consumption 

 was between 35 and 126 gallons per person per 

 day, and when it is considered that one or more 

 of the aqueducts were generally out of service 

 owing to the progress of repairs, the lower 

 figure is probably nearer to the actual consump- 

 tion. Mr. Herschel's final conclusion is that 

 the probable daily consumption was 38 gallons 

 per person, although the actual value doubtless 

 varied some 20 gallons on either side of that 

 figure. 



The Roman laws regarding the injury of 

 aqueducts, the pollution of their waters, and 

 the unlawful use of water have formed the 

 basis of modern statute law for the protection 

 of public water supplies. Regarding the dis- 

 tribution of water to buildings, it is interesting 

 to note that direct connections with the aque- 

 ducts and street mains were forbidden ; these 

 mains delivered the water to small distributing 

 tanks, and a house connection was made to one 

 of these tanks by a quinaria pipe. It was re- 

 quired by law that this pipe could not be in- 

 creased in diameter within a distance of fifty 

 feet from the tank, since by so doing the dis- 

 charge would be increased. These water tanks 

 were under the charge of men called 'aquarii,' 

 who probably bought the water from the city 

 and sold it to consumers, since various methods 



devised by them to defraud both the city and 

 the consumers are described and severely de- 

 nounced by Frontinus. 



Roman arithmetic and mensuration form the 

 subject of one of Mr. Herschel's interesting 

 chapters, but his statement that Frontinus used 

 for TT the value 31 seems to be scarcely war- 

 ranted, la fact the list of fractions used by 

 Frontinus does not contain \, that list being J, 

 ii ii i> T2i 5ii 7i;i «i yVi jJ^) and multiples of 

 these. From the diameters and circumfer- 

 ences of several pipes, as stated in De Aquis, 

 the writer concludes that the value of tt used in 

 computing the latter from the former was prob- 

 ably S/sV, or, in the notation of Frontinus, 



3 4- T3 + 2V + tV + 5^S"- 



Mr. Herschel remarks that Frontinus takes 

 high rank, even to-day, as a practical superin- 

 tendent of waterworks. His first care on as- 

 suming the oflice was to ascertain the history of 

 the aqueducts and the regulations and customs 

 regarding the measurement of the water ; then 

 he made gaugings to ascertain the amount of 

 water carried by the aqueducts and the con- 

 sumption. Finding the latter far less than the 

 former, he traced the discrepancy to unlawful 

 connections. The laws relating to these were 

 then enforced and measures taken to prevent 

 waste. The result was that " the cleanliness of 

 the city is greatly improved, the air is purer, 

 and the causes of pestilence are removed"; at 

 the same time "those who with fear drew 

 water unlawfully, draw their supply now free- 

 from care, by grant from the sovereign," and 

 thus morality was promoted. 



It is a matter of congratulation that the first 

 English translation of Frontinus' 'De Aquis 

 should have been made by a hydraulic engineer 

 highly qualified to interpret it. In fact our 

 knowledge of Frontinus and his work is mostly 

 due, not to classical scholars, but to technical 

 men, the French and German translations hav- 

 ing been made by the civil engineers Rondelet 

 and Dederich. The writer of a history of Rome, 

 published at Boston in 1886, refers to the book 

 of Frontinus as belonging " to the class which 

 furnish facts without giving ideas." If, how- 

 ever, Frontinus be read carefully by those who 

 are able to understand the facts, as scientists 

 and civil engineers can do, it appears, as the- 



