~ 
TRANSACTIONS OF SECTION G. 655 
abandonment of that method by one or two of the highest authorities and most 
extensive experimenters. . These difficulties fall under two heads :— 
1. The construction of a suitable meter and the determination of its constants. 
2. The mode of using it to obtain sub-surface velocities. 
The meters which are by far the most generally used have a revolving screw or 
fan, the number of turns of which in a given time affords a measure of the speed 
of the current. Instruments of this class have been brought to a tolerable state of 
perfection, and by means of various devices by which electric communication is 
established between the screw and the observer at the surface very satisfactory 
results have been attained. The mode of using this kind of meter at comparatively 
small depths and moderate velocities is to employ a rod of wood or metal, or an 
iron tube, by which it is held in the required position. Where the channel is deep 
and the current swift this method requires either elaborate raft or other arrange- 
ments, or the assistance of several men. If, as is very often the case, the channel 
is a tidal and navigable one, and interruptions are frequent, the taking of a series 
of observations by these means is a toilsome and laborious task. It is under the 
latter conditions that the author is at present engaged in taking a series of obser- 
vations, and this paper contains a brief account of certain instruments employed, 
and the mode of using them. 
One object of the experiments was to obtain velocities at one point near the 
bottom during the whole rise and fall of the tide. To avoid the labour of frequent 
observations and the continued attendance above the point with a boat, which 
would have been otherwise necessary, the plan was tried of supporting the meter 
at the bottom of the channel instead of suspending it from above. This was done 
by driving an iron bar into the river bed at low water, and screwing the meter to 
it in its right position, and at such a depth as to avoid all danger to or from 
passing shipping. 
A self-recording meter was necessary, and the one exhibited was employed. 
The instrument in its original form has been elsewhere described by the author, 
but the present form has a most important modification in the recording appa- 
ratus. Into the water-tight barrel a spindle passes, which is turned once for 
fifty revolutions of the screw. At every revolution of the spindle a needle is 
raised which flies back under the action of a spring and punctures a tinfoil 
sheet, a specimen of which was shown. This sheet is wrapped round a drum, 
which is turned uniformly by clockwork once in an hour, so that not only can 
the velocity be determined by the number of dots in a given space, but also the 
time at which the particular velocity occurred. In order to use the whole surface 
of the foil, the piece which carries the marking needle and spring moves along by 
a slow screw, so that the instrument will record continuously for as long as 
twelve hours. There is yet another point, viz., that in order to record the time 
when the tide turns another needle is used, which records the motion in the 
opposite direction, and the marks on the foil slant the opposite way. 
The special objection which was found to hold with this arrangement is the 
fact that weeds and drift moving along near the bottom get entangled in the fan. 
In this case there is nothing in the record to enable the cause of the consequent 
alteration to be detected with certainty. Moreover, the bearings of the fan need 
frequent examination in the muddy waters in which the instrument was employed, 
and this it was, of course, impossible to secure. 
For this reason the author afterwards employed at all depths the method, 
which has recently been used by Professor teers and other observers, of sus- 
pending the meter with a weight. A suitable tail causes the meter to take its 
proper position in the current, and.it can be easily hauled up for inspection. 
The author was led to conclude that for the above reasons no meter with a 
revolving screw could be left to work by itself at any rate under these conditions, 
Also that the method of suspension by a rope is by far the easiest and most rapid 
mode of experimenting. 
There are, however, at least three serious objections inherent in screw current 
meters: 
' Minutes of Proc. Inst. C. H., vol. 1xix. Pp. 3999 
