6 H. J. Harman—Operations for obtaining [No. 1, 
of them remained above water (fig. 8). In the 
mouth of the tube was a cork and some cotton 
wool for a marker. I obtained from Calcutta a 
number of hollow balls 3 in. diameter, for sub- 
surface velocity measurements: I did not get them 
till February and found that unfortunately they 
were defective and were of little use to me. 
Thad also a current-meter by “ Casella,” with 
which I intended observing sub-surface velocities. 
Capt. Willans, R. E., Executive Engineer, Shil- 
long, very kindly designed and had made up for 
me an effective apparatus, a lift by which the 
Fig. 3. current meter could be lowered and fixed to any 
desired depth in the water, could be started and stopped at pleasure, and 
raised easily to allow of the record being read off. Unfortunately, when 
too late, I found that the meter would not work except with a very high 
velocity of water, and those records obtained were untrustworthy. 
The method of measuring the velocity was as follows :— 
The boat was moored on the upper line A C (vide Fig. 1), and the 
floating instruments dropped into the water. The general line of direc- 
tion taken by the floats was observed with a prismatic compass ; on the bank 
were two observers M and N, each furnished with a good pair of binoculars. 
A recorder, with a large chronometer (Dent’s) beating half seconds, was 
seated at O. 
The flag staves were thin and straight. When the observer M saw a 
float cross his line H F he cried “past,” and the recorder at O noted the 
time in his book: when the float passed the line G H, the observer WV cried 
“past” and the recorder noted the time. The interval between the two 
noted times, gives the time taken by the float to pass over the “ Run.” 
The following example shows the method of computation adopted in 
the majority of cases: of course it sometimes happened that the section of 
the bed of the river between adjacent sections where velocities were taken 
Sra Level of water S”b was as in fig. 4, or the direction of 
é = the flow of the water at a differed 
== greatly from that at b, or the obser- 
oe vations at one of the Sections were 
Picola untrustworthy ; in such cases modifi- 
cations were egescary and methods recommended by one’s judgment 
were resorted to; but the following example gives the type of computation 
adopted in most cases :-— 
The length of run = 50 feet. 
