ff rOBK MAM Ml -I I M 



iteriala used in the actual man- 



certain in some caaes, be- 



burning, it ia a decided defect in the 



guarded againsl by the presenl Bpecificationa. 



ed above would go Ear toward pre- 



ceo f such cements, but a more direct method 



... result would seem to be preferable. This 



; lt . be doi changing the last two paragraphs of § 35 of 



pecificatioi to read: 



,1 f three parts by weight of standard crushed 



quart/, and one pari of weight of Portland cement, mixed in the 



manner and kepi Beven days under the same conditions, 



il show a Least average tensile strength of 125 lb per square 



inch. 



BriqTj parts by weight of standard crushed quartz 



one part by weight of Portland cement, mixed in the same 



,,,„.,. and kept 28 days under the same conditions, shall show 



, ast average tensile strength which shall he at least 50# greater 



than that Bhown by briquets from the same lot, tested at seven 



davs, and Bhall in uo case he less than 220 lb per square inch. 



Results of the tests 



The results obtained during the four years noted have been 

 tabulated in the following pages. A few tests have been omitted, 

 when- the circumstances seemed to render it advisable to take this 



ion. 



On pi. 1-12 are Bhown the average results obtained in the tests 

 for fineness, timeof Betting, and tensile strength, the cements 

 having been grouped according to locality. From these diagrams 

 ; t v ,j]; the \ew York Portlands have for three years 



out of the four given the highest results for fineness; that in 

 t( ., rength they are aboul on an equality with those manu- 



factured in Pennsylvania, the product of both states averaging 

 than do the two New Jersey brands; while the results 

 ts for time of Betting are the most variable. 



