74 



HYDROGRAPHICAL SURVEYING [chap. hi. 



Signal to 

 be made. 



Watch to 

 be used. 



Prepara- 

 tion for 

 counting. 



Repeti- 

 tions. 



Calculat- 

 ing the 

 Mean. 



of the wind as possible, as it is the whistHng of it in the ears 

 wliich disturbs the receiver more than anything. 



A base of 3 miles is a very good length, but the surveyor 

 will generally not have much choice in this matter. Needless 

 to say, on a calm day the sound will be heard farthest and 

 easiest, but the choice of days is seldom possible in practice. 

 If we waited for the best opportunity for every detail of 

 survey, it would never get on, and the utmost that can be done 

 is, when there is alternative work for which the day or oppor- 

 tunity is more suited, to take that in hand. 



The guns from the two ends should be fired alternately, at 

 regular intervals, and at some preconcerted signal, as dipping 

 from the ship a flag visible from both stations, which should 

 be hoisted a minute or half a minute before as warning, or re- 

 hoisting a dipped flag steadily, the gun being fired as the flag 

 reaches the mast-head. It is distracting to the receiver to be 

 waiting an indefinite period for the flash. 



A chronometer watch is the best, beating five ticks to the 

 two seconds. An ordinary watch, which beats nine ticks in 

 the same period, goes at such a pace as to be rather confusing, 

 especially when not in practice, though, if the observer is used 

 •to the process, he will measure as accurately with an ordinary 

 watch, and possibly more so. 



When awaiting the flash, hold the watch to the ear and 

 count to yourself — nought, nought, nought, etc., continually, 

 keeping time with the ticks ; you A\ill then be ready to com- 

 mence — one, two, three, etc., as soon as you see the flash or 

 smoke of the gim. 



If going to use a telescope to watch for the warning signal, 

 tie the watch over the ear ^\^tll a handkerchief, which will leave 

 both hands free. 



Count only up to ten or twenty, and mark off each ten or 

 twenty by putting down a finger of the unoccupied hand, or 

 by some such means. 



If time allows, three or four measurements should be made 

 each way, or more if they do not agree with one another. A 

 signal must be arranged to ask for more than the number 

 previously settled, if it be wanted. 



In meaning the result, the arithmetical mean is not strictly 

 correct, as the acceleration caused by travelHng with the wind 



