584 EXPEEIMEXT STATTOX RECORD. 



but they are not so dependable for continuous operation as steam engines, and 

 the latter equipment is, therefore, generally preferred for large machines. . . . 



" If a landowner expects to install 100 rods of tile drain in soil which would 

 require picking, but which contains no rock, he would be justified in buying a 

 ditching plow costing as much as .?20. If he proposes to construct 1,500 rods 

 of tile drain in soil free from rock and large roots, the landowner can well 

 afford to purchase a horse-drawn ditching plow costing from $250 to $300 ; and 

 if it be assumed that the owner can sell his machine when his ditching is com- 

 pleted for $100, he would be justified in purchasing such a machine for the 

 construction of 1,000 rods of drain. For the installation of as much as 5,000 

 rods of drain in a soil free from rock and large roots, the purchase of a power- 

 driven trenching machine costing as much as §1,500 probably would be justified, 

 on the assumption that the machine could subsequently be sold for one-half 

 its original cost. ... In comparing the real costs of differejit machines and 

 implements one must consider not only the purchase price and the operating 

 cost of fuel, oil, and labor, but also repairs, interest on the investment, and 

 depreciation." 



Investigation of the durability of cement drain tile in alkali soils, R. J. 

 Wig, G. M. Williams, et al. {V. S. Dept. Com., Bur. Standards Technol. Paper 

 U (1915), pp. 56, pis. 20, figs. 16; abs. in Engiti. Rec, 12 (1915), A'o. 8, p. 220).— 

 This bulletin describes several concrete structures exposed to alkali water which 

 have failed in the Western States, and reports the first-year results of service 

 and laboratory tests of 8,800 hand and machine made concrete drain tile made up 

 of 16 different types, the proportions of cement and sand varying from 1 : li to 

 1: 4. The curing was either by sprinkling or by steam. The tile were installed 

 in operating drains on eight projects, comprising the most concentratetl alkali 

 soils of the West, and for comparison on projects where there was practically 

 no alkali. 



The results of the tests of tile were, as a whole, quite uniform and compared 

 favorably with other tests of drain tile. Occasionally results of tests of similar 

 tile differed by 30 or 40 per cent, but usually they agi-eed within a few per 

 cent. Some of those which had a normal or only slightly reduced strength 

 showed evidence of alkali action by softened edges or cracked surfaces. " The 

 great majority of the tile were unaffected in any manner at any of the projects. 

 At Sunnyside, Wash. ; Yuma, Ariz. ; and Roswell, N. Mes., none of the tile 

 showed any abnormality whatever. At Garland, Wye, the leanest mixtures 

 showed reductions in strength considerably below the averages for these series 

 at all points." At Fort Shaw, Jlont., reduced strength was shown in two cases 

 of handmade, steam-cured tile of 1 : 2 mixture of plastic consistency. "At Grand 

 Junction, Colo., the tile made of the leanest mixture were swollen and cracked. 

 The tile made of sand-cement, while of normal strength, showed a softening at 

 the edges, and several square inches of one of the tile were similarly affected, 

 although the breaking strength was normal. At Huntington, Utah, the tile of 

 series 6 [handmade ,steam-cured, 1 : 2 mixture] appeared to be affected by the 

 alkali. Although these tile withstood loads of over 1,400 lbs., there were a 

 number of longitudinal and circumferential cracks apparent before testing, 

 and the fractured surfaces appeared white. ... At Montrose, Colo., the tile 

 [made of sand-cement] appeared to have been affected by alkali. On one of 

 the tile the concrete at one end was softened, apparently due to alkali, and 

 both failed under comparatively small loads. . . . 



" The absorption of moisture by the walls of the tile of the various series 

 was noticeable at the different projects. The apparent amount of absorption 

 varied with the amount of water present in the soils at the various points, and 

 it was found that some of the series wouUl be apparently saturate<l where the 



