TRANSACTIONS OF THE SECTIONS. 19 



retardation : and in a circulating stream the quantity of etiergy ofjiow gained or lost in 

 each complete circuit is equal to the quantity/ of energy lost or gained in the form of 

 heat; and in the absence of friction, the ratios home hy that quantity to the heat added 

 and the heat abstracted (of which it is the difference) are regidated hy the ahsolute 

 temperatures at which heat is added and abstracted, agreeably to the second late of 

 thermodynamics. 



Amougst particular cases of the thermodynamic acceleration and retardation of 

 streams, the following may be specified. 



Acceleration by the addition of heat at and near a place of maximum pressm-e : — 

 the di-aught of a furnace ; and the production of disturbances in the atmosphere in 

 regions where the ground is hotter than the air. 



Retardation by the abstraction of heat at and near a place of maximum pressure : — 

 the dying away of atmospheric disturbances in regions where the gi-ound is colder 

 than the air. 



Acceleration by the abstraction of heat at and near a place of minimum pressure : — 

 the injector for feeding boilers, in which a jet of steam, being liquefied by the 

 abstraction of heat, is enabled not only to force its way back into the boiler, but to 

 sweep a current of additional water along with it ; also, to a certain extent, the 

 ejector-condenser. 



The conduction of heat from the parts of a stream where the pressure and tem- 

 perature are highest to the parts of the same stream where the pressuj^e and tempera- 

 tm"e are lowest, produces, according to the foregoing principles, a gradual and per- 

 manent retardation of the stream, independently of the agency of friction ; and this 

 is accompanied by the production of heat to an amount equivalent to the lost 

 energy of flow. 



Report of the Liverpool Compass Committee. By John T. Towson. 



The last Report read to the British Association was in 1859, at Aberdeen, The 

 most important result that had occm'red since that time was the cessation of the 

 difference of opinion that liad previouslj' existed between those connected with the 

 royal navy and with the mercantile marine, the former advocating the use of a 

 table of errors, the latter mechanical compensation on the principle introduced by 

 the Astronomer Royal. In 1854:, Mr. Towson said his advocacy of compensation 

 was scarcely tolerated by some Members of the Association; but in 1869, Mr. 

 Archibald Smith, one of the Authors of the 'Admiralty Manual on Compass 

 Matters,' stated before the Royal Society that the question of mechanical compen- 

 sation of the compass had materially changed in its aspect of late years, and he 

 advocates its use in most cases. The next matter was of minor importance, but 

 proves that neither a table of errors nor mechanical compensation can be relied on 

 within the limit of three degi'ees. If the ship be swimg to the north or south, 

 from left to right, the needle wiU be drawn three degi-ees more to the right than it 

 would be if the ship were swung in the contrary direction. The most valuable 

 result which has been brought about within the last year is the establishment of a 

 voluntary examination of masters of iron ships in the theory and practice of com- 

 pass-deviation. 



On Non-tidal Variations of the Sea-level on the Coast of India. 



By W. Paekes. 



This paper gave some results based on tidal observations taken at Kurrachee, 

 near the mouth of the Indus. It described a graphical process for eliminating the 

 changes of sea-level due to semidiurnal and diumal xmdulations, thus exhibiting 

 only those which are due to tides of long period or other causes. Referring to the 

 diagram for the month of November 1868, which is appended to the Report of the 

 Tidal Committee in this volume, it will be seen that the black dots representing the 

 successive heights of high and low water range themselves in four waving lines 

 along the paper. The line which represents " half-tide level " was so drawn as to 

 be the locus of a series of ordinates, each of which is a mean of the corresponding 

 ordinates of these four lines. 



By measuring off the ordinate (or height above a fixed level) for noon of each 



2* 



