RURAL ENGINEERING. 483 



reporled. the test specimens consisting of 100 eacli of 8-in., IfJ-in., and 24-in. 

 tile of each material. 



The general conclusions drawn are that the results of the tests by each of the 

 tyjies of bearings are consistent and concordant, and apparently reliable and 

 tXHily indicative of the quality of the tile. The numerical results by the different 

 bearings have fairly uniform ratios and can be calculated and reported in sub- 

 stantially the .same unit by multiplying the brealiing loads per lineal foot by 

 the following factors: Sand bearings, 1; hydraulic bearings, 1.25; and three- 

 point bearings, 1.5. 



The sand bearings required from one to seven minutes more per test than 

 the other bearings but are said to come much nearer to giving the real support- 

 ing strength of the tile in the ditch. The hydraulic bearings distributed the 

 load very well along the tile to fit irregularities in shape and permitted rapid 

 testing. The three-point bearings are considered the simplest and most con- 

 venient of all to use. 



" There can not be nearly so wide a variation of the * ordinary supporting 

 strength ' of drain tile in ditches, to carry the loads from the ditch filling as 

 has heretofore been very generally assumed. ... A comparison of the loads 

 from ditch filling with the results of laboratory tests . . . indicates that the 

 ' ordinary supporting strength ' of drain tile in ditches is approximately equal 

 to the bre.-iking loads in tests with sand bearings." Additional tests of the 

 same nature are reported which confirm the above conclusions. 



A comparison of actually weighed values of loads on pipes in ditches from 

 the weight of ditch filling with those computed from the formula Wf=CicB' 

 shows a con-espondence between the computed and weighed loads, thus closely 

 checking the correctness of the formula. In this formula Wp^the load on a 

 pipe in a ditch, in pounds per lineal foot, from the weight of ditch filling, C=the 

 coefficient, taken from a proper table or diagram, of loads on pipes in ditches 

 from ditch filling. ?t-=weight of ditch filling material in pounds per cubic foot, 

 and B = the breadth of the ditch, a little below the top of this pipe, in feet. 



Investigations of factors of safety in actual tile drains and pipe sewers where 

 the pipe have actually been observed to be sound resulted in the conclusion that 

 with so-called '* first class " pipe laying conditions, corresponding to the best 

 pipe laying practice, and watched constantly by an inspector, it will be safe to 

 use a nominal factor of safety of 1.25. For " ordinary " pipe laying conditions 

 the factor of safety should be 1.5. 



Other sections give data on the manufacture of the concrete tile tested and 



on the calculation of the modulus of rupture of the material of the tile shells. 



W 

 The formulas resulting from the latter calculation are: .li=0.20/2r^, and 



ft \f 



p=--^, where M=the maximum bending moment in the tile shell, in inch- 

 pounds per lineal inch, R=the radius of the center line of the tile shell, in 

 inches, W= the " ordinary supporting strength " of the tile, in pounds per 

 lineal foot, calculated by multiplying the breaking loads in strength tests by 

 the factors noted above (five-eighths the weight of the tile per lineal foot for 

 sand bearings, or three-fourths for hydraulic or three-point bearings, must be 

 added to IF in computing M whenever such addition exceeds 5 per cent of IF), 

 p= the modulus of rupture of the material of the tile shell, in pounds per 

 square inch, and i=the average thickness of the tile shell, in inches, at the 

 top or the bottom, whichever averages thinner. 



Tile investigations, W. H. Day (Ann. Rpt. Ontario Agr. Col. and Expt. 

 Farm, 39 {1013), pp. 56-6.^, figs. 5).— Comparative tests of the breaking strength 

 of 3 and 4-in. cement and clay tile showed that the average breaking strength 



