484 EXPERIMENT STATION RECORD. 



of (ho cement was JO lbs. more (lian tliiit ol' thin day tile, and .'5I>;; lbs. less 

 than average clay tile. It is concluded, however, tliat a\iy of the clay or cement 

 tile tested were stronj? enough to resist the earth pressure to which they would 

 be subjected if placed iu the ground. Further comparative tests of machine- 

 made cement tile and clay tile led to the conclusion that cement tile, if prop- 

 erly made and cured, are as strong as some well-known and satisfactory makes 

 of clay tile and strong enough for tile drainage purposes. 



Tests of the immersio)i iu air, Immersion in vacuum, and si>ecific gravity 

 methods for determining the i)orosity of tile showed that the vacuum and 

 si)ecific gravity methods give almost identical results, which are both more 

 than one-lifth higher than by inunersion in free atmosphere. Using the im- 

 mersion iu vacuum method, it was found that the iwrosity of well-made cement 

 tile w^as on the average considerably less than that of clay tile. It was also 

 found that porosity of the tile wall was not the governing factor in permeabil- 

 ity, this depending more on the glaze of the surface. 



Tests of the permeability of cement and clay tile showed the clay tile, al- 

 though more porous than the cement tile, to be usually less permeable. The 

 cement tile, however, showed considerable difterence of behavior which de- 

 pended on the wetness of the mortar, it being found that if the mortar was wet 

 enough so that the packer produced a smooth watery surface on the entire 

 inner surface the trie wall was practically water-tight. 



Proportioning aggregates for Portland cement concrete, A. Moyeb (Reprint 

 from. AiHcr. Soc. Testing Materials Proc, 14 {I'JlJf), pp. 12). — This paper de- 

 scribes various methods of carrying on investigations so that with a given sand 

 and a given stone or gravel, proportions can be stated by the engineer which 

 will make a concrete of maximum density and maximum strength. Investiga- 

 tion showed that arbitrary specifications without previous knowledge of the 

 character of the aggregates are likely to be wrong. 



It is concluded that it takes 110 lbs. of Portland cement to make 1 cu. ft. of 

 paste in opposition to the usual assumption of 94 lbs. per cubic foot. The au- 

 thor further concludes that instead of the old plan of filling the voids with sand 

 and cement these voids must necessarily be filled with paste (cement and 

 water), and that the study of the proportioning of aggregates must be based 

 on the proper proportions of cement, water, and sand to make a sufficiently 

 rich mortar to bind together the larger aggregates. 



For the purpose of economy it is stated that various sizes of stone should be 

 used and an investigation made to determine which size will produce the least 

 pei'centage of voids so that less mortar may be used and more strength ob- 

 tained. 



Specifications for sand for concrete, E. McCullougii {Cement Era, 12 {Wl.'i), 

 No. 10, p. 56, fig. 1). — The author "as a result of many hundred granulometric 

 analyses of concrete sand proposes the following si>ocificatiou as suitable for 

 general use and which will not bar out any good sand to be found on the 

 market : 



" The sand may be any suitable material of a quality at least equal to the 

 quality of the stone or gravel used in the concrete and shall range in size from 

 fine to coarse. It shall all pass through a l-iu. mesh ; not to exceed SO per cent 

 shall pass through a 20 mesh; not to exceed 50 per cent shall pass through a 

 50 mesh ; and not to exceed 20 per cent shall pass through an SO mesh." 



[Land clearing], J. H. Grisdale {Caimda Expt. Farms Rpts. 1913, pp. 26- 

 2S). — Twenty acres of land covered with heavy timber were cleared under 

 provincial conditions for the plow at an average cost of $506 per acre. The 

 windfalls, trees left by loggers, and brush were cleared away before attacking 

 the standing timber. 



