CONCRETE LINING FOR IRRIGATION CANALS. 53 



In experiments with oil-mixed Portland cement concrete, 1 the con- 

 crete was damp-proofed by the incorporation of semiasphaltic oil. 

 In hand mixing, the sand, cement, and water were first mixed to a 

 mushy consistency, the oil added and mixed until no trace of it was 

 visible on the mortar surface, and then the broken stone or gravel 

 was mixed in. In machine mixing the sand, cement, and water were 

 first mixed to a mortar followed by alternate batches of stone and 

 oil which were added and mixed. The proportion of oil used was 

 based on a comparison of its weight to the weight of the cement used 

 in the concrete. The results of tests show that oil-mixed mortar; 

 containing 5 to 10 per cent of oil, is dampproof as well as water- 

 proof, and indicate that its use may prove desirable in the construc- 

 tion of irrigation canal linings exposed to the action of alkali (fig. 1). 



Good practice in concrete lining construction where alkali must be 

 reckoned with necessitates the following precautions : 



(1) Do not use sand, gravel, or water containing alkali. 



(2) Keep soil waters charged with alkali from coming into contact 

 with the concrete by the use of suitable drainage. 



(3) Give careful attention to the proper proportioning of materials 

 and use more cement than is needed to fill the voids. 



(4) Protect the surface by a thin plaster coat of dense mortar of 

 granular sand. 



(5) Both the concrete and the mortar used for the lining may be 

 dampproof ed by the addition of 5 to 10 per cent of semiasphaltic oil 

 when mixing the materials. 



THE EXPANSION AND CONTRACTION OF CONCRETE. 



FIELD TESTS AT LOGAN, UTAH. 



During the summer of 1913 field experiments were conducted by 

 Prof. B. P. Fleming, of the State University of Iowa, working under 

 the direction of this office, at Logan, Utah, for the purpose of deter- 

 mining the coefficient of expansion of concrete slabs. An effort was 

 made to secure conditions as nearly as possible like those found for 

 canal linings. It will be noted, however, that the slabs were not 

 tested to find the effect produced by being wet on one side. 



The variations in length were measured with two micrometer 

 microscopes focused upon lines in the highly polished tops of two 

 steel pins projecting above the surface of the slab, one at each end. 

 This device permitted making direct readings from the micrometer 

 scale, giving measurements to 0.0008 of a millimeter. The ther- 

 mometers used were graduated to 0.10° C. and could be estimated 

 easily to 0.025°. All measurements of the slab length were direct, 

 and no part of the apparatus was in contact with the slab. The 

 idea of maintaining the latter condition was to prevent in every 

 possible way influences which might affect the expansion of the slab. 



i U. S. Dept. Agr., Office Pub. Roads Bui. 46 (1912). 



