12 DEPAKTMENT BULLETIN lUO. 



16 inch weir, an upper head of 6^ inches and a lower head of 4^ 

 inches. To find the discharge, it is first necessary to convert the 

 heads in inches into feet. This conversion is readily made with the 

 aid of Table 3, which shows that 6^ inches equals 0.542 foot and 4^ 

 inches equals 0.354 foot. Of course, if both heads are measured in 

 feet, no conversion is necessary. In order to find the difference of 

 head, subtract 0.354 from 0.542 and get 0.188 foot, or approximately 

 0.54 and 0.19, respectively, for the upper head and the difference of 

 head. Next turn to Figure 8, which is the discharge diagram for a 

 2 by 16 inch weir 1 foot long. On the left of the diagram, find the 

 horizontal line for the 0.19 difference of head, follow it to the right 

 until the 0.5-upper-head line is reached, estimate the position of the 

 0.54-line, and then find where the 0.19 difference-of-head line cuts the 

 0.54-upper-head line, and directly beneath this point read 0.98 cubic 

 feet per second as the discharge for a 1-foot flume. Then for a 3-foot 

 flume, the discharge is 3 times 0.98 or 2.94 cubic feet per second 

 under the conditions given above. These diagrams should not be 

 used for flumes less than 1 foot nor more than 4 feet in width. 



It will be observed that the diagrams do not cover the complete- 

 range of differences possible for each upper head; in fact, only the 

 small differences are shown except for the small upper heads. This is 

 due to the difficulty in obtaining these heads in practice. Where large 

 flows with large differences in head are measured, the velocity of the 

 water is so great below the crest that there is very little tendency to 

 submerge the weir until the retardation below is quite great. It is 

 doubtful even, if these large differences of head are possible except 

 for very high weirs which tend to reduce the velocity of the water 



below the crest. 



SPECIAL TESTS. 



The results of the special tests to determine the effect of changes 

 in the flume are given in Table 4. The results of these tests show that 

 neither the changes in the crest nor the changes in the approach con- 

 dition cause errors greater than 3.3 per cent. For submerged condi- 

 tions, the errors would probably be less on account of the lower veloci- 

 ties of flow. This indicates that the changes which ordinarily occur 

 in the fleld would not cause serious errors in the measurements of 

 discharges from this type of measuring device. 



