130 Bulletin 86 



put in the line, and the first run of water should fill the pipe without 

 subjecting it to much pressure head. 



The exact relation between the loss of strength and the absorp- 

 tion of water can be obtained only by systematic testing on a com- 

 prehensive plan. The hypothesis is offered, however, that the weak- 

 ening can be reduced by extending the period of curing, and this 

 should be done in most cases. Most pipe makers cure their pipe 

 only from six to ten days, and the duration of time of curing ot 

 large sized pipe or of pipe that is not to be laid immediately may 

 well be extended to 15 or even 30 days. 



External pressure tests were made on another group of pipe on 

 April 27, 1918. The data are given in Table XIII. Pipe No. 1 of 

 this group was a 14-inch pipe, which had been in use in the ground 

 for over a year but was removed during some change in the line. It 

 was dried in the sun before testing. The pipe was under the usual 

 weight because the tongue and groove were broken off. The next 

 three pipe were five months old and were tested dry. 



The 20-inch pipe consisted of four specimens that were made of 

 mortar mixed in the proportions of one of cement to three of sand, 

 and four specimens of 1 :234 mortar. The two lots were approxi- 

 mately the same age. The leaner mixture showed the greater 

 strength. This is contradictory to common knowledge of the varia- 

 tion of strength with richness of the mortar. Possibly it would be 

 better not to publish the results. It may be, however, that these 

 results are linked in some way to the pipe failures described on page 

 110. If differential expansion causes weakening of the pipe, differ- 

 ential contraction at some stage of the curing or just afterward may 

 have caused minute shallow cracks, which would be more likely to 

 come on the outer surface of the pipe, and would tend to weaken the 

 pipe permanently. 



Another test series was made in June, 1918, to determine the ef- 

 fect of variations in the curing of pipe. All of the test specimens 

 were of 16-inch pipe and were taken at the pipe plant at one time, 

 so that they were of the same proportions and consistency. The 

 original program of testing was not carried out, however, be- 

 cause the pipe maker failed to follow instructions for curing the 

 test specimens. 



The results are given in Table XIV. Four pipe were tested at 

 the end of seven days. After 28 days 12 pipe were tested. Of these 

 12, four had been kept wet by sprinkling, four had been cured for 

 10 days and then placed in the stackyard, and four had been cured 

 and stacked but were soaked in water for 5 hours just before test- 



