128 PROCEEDINGS OF THE CANADIAN INSTITUTE. 



A Probable Solution of the Secondary Undulations Found Upon 

 Self-recording Tide Gauges. By Napier Denison, Esq., Toronto 

 Observatory. 



(Read January i6, 1897.) 



As this is the year set apart throughout the scientific world for special cloud 

 'Observation and the study of the upper atmosphere as a means of improving the 

 present methods of weather forecasting, and being one of those engaged upon this 

 work for Canada, it seems an opportune time to bring the following investigations 

 -before the public : 



In June last the writer's attention was first drawn to small, rapid changes of 

 water level on Lake Huron, at the mouth of the river at Kincardine, where a rise 

 and fall appearing to be regular, a set of observations with a temporary float were 

 ■taken, and a uniform rise and fall of about three inches was found to occur, 

 averaging nine minutes, that is about eighteen minutes for each undulation, the 

 float moved up stream at the rate of a mile and one-half an hour. Upon returnmg 

 to Toronto in July, by permission from Mr. Stupart, Director of the Meteorological 

 Service, a simple instrument was devised to automatically record such oscillations, 

 and set up at the mouth of the Humber River, where, ever since, most interesting 

 jresults have been obtained, (i) By comparing these with the sensitive photographic 

 barograph traces of the Observatory, it was found when a rise and fall of about 

 four inches in water level occurred a corresponding, but slight change, in atmos- 

 pheric pressure is noticeable; also a marked twenty-minute interval between these 

 waves. To fully investigate this phenomenon another instrument, similar to that 

 at the Humber, was placed at the Burlington Canal, and records from both 

 instruments plotted, in conjunction with the- barograph traces and different types 

 of weather, as taken from the bi-daily weather charts. 



It has been found, ever since the introduction of self-recording tide gauges 

 throughout the world, peculiar oscillations within the normal tidal curves have been 

 noted and commented upon, but to my knowledge, the cause not fully accounted 

 for. These oscillations are most marked at tidal stations situated near the ends of 

 bays, owing to the favourable configuration of the shore, and the shallowness of 

 the water. They are not, liowever, of constant occurrence, but vary in intensity 

 upon different days. Some months ago my attention was drawn to this peculiar 

 -phenomenon by a paper read before the Royal Society of Canada in May, 1895, by 

 W. Bell Dawson, Ma.E., C.E., in charge Canadian Tidal Service, (2) who graphically 

 demonstrated the existence of these secondary undulations, but left their cause 

 unexplained. Thinking the above lake researches might help to solve the cause of 

 these secondary tidal curves, through the kindness of Mr. Dawson, copies of the 

 -tidal records for Halifax, N.S., and St. John, N.B., were obtained for a number of 

 days, including different types of weather. To fully investigate all the phenonema 

 pertaining to these curves, days in succession were taken, chiefly those precedmg 

 and during stormy weather, for instance, the 5th, 6th, and 7th of February, 1896. 

 These curves were plotted upon one-tenth inch squared paper, allowing one mch to 

 -the foot for Halifax, which has a range of about six feet, and one inch to five feet 

 for St. John, which has a range of about twenty-eight feet, (3) and allowmg one- 

 tenth for every five minutes, being careful to minutely measure the extent and 

 duration of every oscillation; above this was plotted the readings of the Halifax 

 barometer, taken every three hours, also the hourly direction and velocity of the 

 wind. 



(i) A full account of these lake researches to be published shortly. 



(2) Note on secondary undulations recorded by self-registering tide gauges. Vol. I., Sec. III. 

 ,) The St. John tidal curve is not included in above table as only the most pronounced undulations 

 are noticeable due to the reduced scale necessary where such high tides occur. 



