398 Mr. A. H. Gibson on the Resistance to 



justified, the logarithmic homologue for each pipe will be a 

 straight line; and these straight lines will all pass through 

 the same point, the abscissa of which represents log B while 

 its ordinate represents log A. The plotted results afford 

 ample justification of the assumptions ; for certainly within 

 a limit commensurate in magnitude with the limit of accuracy 

 in the experiments, the plotted lines lie on a series of straight 

 lines all of which pass through the point (antilog 6*097, 

 antilog-8200). 



• This makes A = 125 x lO" 8 , B = 6'60; so that equation (6) 

 becomes 



8i? = -Q000012.5 j ^~ 1 ' t? » t ' Z lbs. per so. in. . . (7') 

 6-6 n .d 3 ~ n 



The following are the values of 'ra' as deduced from these 

 experiments. 



f Lead pipe "125 in. diam. Very smooth inner surface. n 



I Author 1-28 



<J „ „ *250 in. diam. Surface condition unknown, 

 I probably moderately rough. 



I I3rix 1-49 



J Cast Iron pipe 7*87 in. diam Stockalper .... 1*775 



1 n „ „ 11*75 in. diam Brix 1*77 



f Wrought Iron pipe 3*25 in. diam. . . Brix 1*825 



J „ „ ,, 5*2 in. diam. . . Brix 1*81 



j „ „ „ 5*9 in. diam. .. Stockalper .... 1*81 



L „ , t » 6 '? 5 in - diam - • • Brix 1 ' 79 



As might be expected, with fairly large diameters the 

 condition of the pipe surface only appears to affect the 

 value of i n 9 slightly, this value appearing (as might 

 also be expected) to diminish slightly with an increasing 

 diameter. 



It would appear from these results, that for either cast- 

 iron or wrought-iron pipes laid under normal conditions as 

 to jointing, &c, n may be taken as having the following 

 mean values. 



Diam. (ins.). 



3. 



5. 



7. 



9. 



12. 



n 



1*83 



1-81 



1*79 



1*775 



1*77 





while for all cases of flow through such pipes, where the 

 air is at atmospheric temperature (about 66° F.) the drop in 



