XLVI. J. I. HAYCROFT. 



or ' 3 panel trestle,' as may appear to his judgment right, and subsequent 

 experience generally proves that he was wrong." 



Mr. Foster Crowell states : — 



" There are all sorts of conditions requiring all sorts of provisions, in 

 the matter of waterway design ; in unsettled districts the flood marks are as 

 a rule more easily discernible than in the settled districts, and where the 

 rainfall is known approximately, and the watershed can be conservatively 

 estimated, even the absence of flood marks, itself a reassuring circumstance, 

 need not preclude a correct diagnosis. A tentative treatment is, however, 

 often resorted to in the adoption of a surplusage of opening, to be afterwards 

 curtailed in the light of experience. In this country (America) reliable 

 records are often lacking, even in settled communities, and where the 

 consequences of erroneous conclusions would be most serious, thorough 

 investigation of the possible discharge becomes a necessity. In the matter of 

 culverts, the usual practice is extremely conservative, especially as under high 

 embankments actual economy may be secured by an increase of diameter of 

 the opening. The question of waterways should not be left until the time of 

 actual construction, but be taken up with the preliminary surveys and in 

 detail. It is, however, not good practice to steer very close to the theoretical 

 requirements. Ample margin should always be allowed where practicable." 



From other letters received by the writer from America and 

 Canada, the practice seems to be, when a doubt exists as to the 

 reliability of the result of a formula, a trestle is put in, about 

 the sufficiency of which there can be no doubt, and this structure 

 is observed in flood times, during a course of years, and when 

 needing replacement is substituted by a permanent culvert, the 

 size of which is determined by the observations made during the 

 life of the trestle. 



The formula most used in America seems to be that of E. T. D. 

 Myers, President, Richmond, Fredricksburg and Potomac Pail- 

 way. Mr. Myers states " that the co-efficient should be derived 

 from careful and judicious gaugings, at characteristic points 

 within the region under treatment, and applied with a liberal 

 hand." 



A valuable paper on " How to determine size and capacity of 

 openings for waterways" has been compiled and published by the 

 committee of the Pailway Superintendents of Bridges and 

 Buildings, in America, in which it is stated, inter alia, after 



