_56 PROCEEDINGS OF THE CANADIAN INSTITUTE. 



ly, which by means of clockwork, completes one revolution every 24 hours, the hour 

 intervals being, therefore, each one inch. Resting upon this cylinder is a self-inking 

 pen, attached to an arm, which slides freely upon a horizontal tubular guide. To one 

 side of this arm is fastened a silk line, which is attached to and wound several 

 times around a small, grooved pulley, which is part of onp four times its diameter. 

 Upon the grooved circumference of the larger pulley is fastened another line, which, 

 after several turns around it, passes down through the case to a float, enclosed in 

 a special shaft, so constructed as to admit the water only through several small holes, 

 and thus to prevent any sudden movement of the float being caused by local wave 

 motion. To the other end of the sliding arm is fastened another line, which, after 

 passing over a small grooved pulley (shown at the left of the illustration), descends 

 through the case, and has attached to it a weight sufficient to balance the float. The 

 record is obtained in the following manner : As the float rises and falls, the pen 

 correspondingly moves up and down upon the paper, which is revolving at the 

 rate of one inch per hour. 



The ratio of movement between pen and float is as i to 4, so that an actual 

 rise of one inch of water level corresponds to a movement of one-fourth of an inch 

 upon the paper on the cylinder. To facilitate the measuring of these traces, one- 

 quarter inch squared paper is used, the vertical lines marking fifteen-minute inter- 

 vals, and the horizontal, one inch change in water level. To prevent the water from 

 freezing in the shaft during the winter months, oil was used, which had the effect 

 of depressing the water level below the frost line. 



In order to increase the value of these records, a similar instrument was set up 

 at the Burlington Canal last September. Before bringing before you some inter- 

 esting tracings taken from these instruments, permit me to summarise previous in- 

 vestigations in other countries. 



This phenomenon had been noted by Duillier as early as 1730, upon the Swiss 

 lakes, where it obtained the name of seiche, owing to the apparent " drying up " or 

 recession of the water upon one side of the lake, when rising at the other. In 1779 

 De Saussure remarks that he believes local variations in the air pressure may be the 

 cause. In 1804 Vaucher published his researches on the subject. His conclusions 

 are briefly as follows : — 



(i) Seiches more or less considerable occur in all lakes. 



(2) They occur at all seasons and at all hours, but most frequently in spring 

 and autumn. 



(3) The condition of the atmosphere is the governing cause ; the more settled 

 this is the less are the seiches, and the more variable it is • the more marked 

 are they. 



(4) Although most frequent in spring and autumn, the greatest oscillations 

 always occurred in July, August, or early in September. 



(5) Although the duration of these seiches is extremely variable, their intervals 

 do not exceed 20 to 25 minutes, and are frequently less. 



(6) They varied much in amplitude at different points on the lake shores, being 

 on the Lake of Geneva greatest at the mouth of the Rhone. 



Vaucher supposed that the atmospheric pressure diminished over one part of 

 the lake, while over another it remained constant or increased. If this change in 

 pressure, occurred suddenly, the water which had thereby been set in motion would 

 not come to rest again until after a number of oscillations. 



Professor Forel agrees with this theory, which has also been accepted by 

 Studer, Meyer and Favre. From 1854 to 1856 an important series of observations 

 were made by six observers, placed at different points on the shore of Lake 

 Geneva, who, using a system of signals to warn each other of the approach of an 

 oscillation, noted the variations of the barometer and of the lake level. As Professor 

 Forel in his article entirely disregards these barometric observations, they do not 

 appear to have been published. In 1870 Professor Forel studied the seiches at the 



