130 A. W. Duff — Secondary Undulations 



(1). It lias been supposed that the results are due in reality 

 to peculiarities in the action of each tide gauge. — This sugges- 

 tion hardly deserves serious consideration, and is only included 

 here because it is so admirably disposed of by the case of 

 Pictou (stated above), where two entirely different gauges in 

 different years showed secondary undulations of the same 

 mean period. 



(2). Earthquake theory. — This also may be dismissed briefly. 

 It could not account for the period nor for the fact that, in 

 such a comparatively small area as the Gulf of St. Lawrence 

 or the Bay of Fundy, some gauges always show secondary 

 undulations, others never show any. 



(3). "Seiche'" theory. — Four years ago I advanced the sug- 

 gestion* that the undulations were of the same nature as the 

 " seiche " movements observed by Forel and others on the 

 Swiss lakes, that is undulations of a partially confined body of 

 .water under the force of gravity, like the " wish-wash " of 

 water in a bowl when disturbed. This explanation seemed to 

 fit entirely the oscillations at St. John (both harbor and river), 

 calculations from the soundings in the vicinity showing a 

 remarkable agreement of calculated and observed period on the 

 supposition that the oscillations were not of the fundamental 

 or gravest mode (uninodal), but of the first higher mode 

 (binodal). As will be seen later, it does not accord so well 

 with other cases. 



It is to be noted that the above theory is chiefly concerned 

 with the period of undulation, leaving the initial disturbance 

 of equilibrium to be accounted for by atmospheric conditions 

 or by neighbouring oceanic disturbances invading the basin of 

 oscillation. 



(4). "Atmospheric billows" theory. — Mr. F. Napier Deni- 

 son has advanced the interesting and ingenious theory f that 

 "the secondary undulations are due to atmospheric waves or 

 billows set up in the upper atmosphere" at the boundary be- 

 tween " the upper part of the lower stratum of air and the 

 higher stratum which is less dense and rapidly moving pole- 

 wards." To account for " how atmospheric waves, which cor- 

 respond to only a few hundredths or thousandths of the 

 barometric inch at the earth's surface, cause such rapid and 

 extreme oscillations" of the water surface, Mr. Denison adds 

 that " the atmospheric waves or billows, in passing over the 



* This Journal, Hi, 1897, where a full account of the bibliography of seiches 

 will be found. In the paper here referred to, following the analogy of sound 

 waves I used the word node (in binodal, trinodal, etc.), as meaning a place of 

 minimum longitudinal motion. Finding it customary to use node in reference to 

 vertical movement when speaking of water waves, I have changed the notation 

 in the present paper. 



f Articles already cited. 



