BAROMETER. 



115 



Combining the two actions, it is seen that 

 from 60 to 90, and from 240 to 270, both 

 waves are descending, while from 150 to 180, 

 and from 330 to 360, both are ascending. 

 Accordingly, besides the principal lunar max- 

 ima and minima at the syzygies and quadra- 

 tures, there should be secondary maxima and 

 minima also at 60 in advance of those points ; 

 and these inferences also the St. Helena obser- 

 vations confirm. 



At extra-tropical stations, Mr. Chase would 

 expect important modifications of the results 

 now given. . Finally, in thus determining the 

 phenomena of aerial tides, he is led to believe 

 that " the long-suspected obedience of the prin- 

 cipal meteorological changes to fixed mathemat- 

 ical laws is at length demonstrated." Proc. 

 Amer. Philos. Soc., vol.-ix., p. 395, as quoted in 

 Amer. Jour, of /Science. 



The Principal Causes of Barometric Fluctu- 

 ations. From this paper, which is occupied 

 with a summing-up and generalizing of the re- 

 sults of the several discussions thus far pre- 

 sented, our extracts must be brief. The author 

 concludes that there are four important causes 

 of barometric disturbance : 1, rotation, with its 

 quarter-daily phases of alternate aid and op- 

 position to the attraction and temperature 

 currents, and of shifting the aerial particles to 

 levels of greater or less density ; 2, variations 

 of temperature and vapor ; 3, lunar attraction ; 

 4, solar attraction. Among subordinate causes, 

 he would reckon, 5, resistance of ether. The 

 influences of rotation and attraction can be cal- 

 culated. Through averaging a long series of 

 hourly observations, the effects of lunar attrac- 

 tion may be so nearly eliminated as to show 

 the approximate value of the other principal 

 disturbances. The formula for the rotation 

 tide is given in one of the preceding papers. 

 Next in order of importance are the tempera- 

 ture and vapor tide, and the solar tide. It is 

 scarcely possible to fix at present the precise 

 amount of disturbance attributable to each of 

 these latter; but the following considerations 

 appear to lead to probable results. 



The theoretical maxima of the rotation-tide, 

 allowing an hour for inertia, occur at 4 h and 

 16 11 ; the minima at 10 h and 22 h . Allowing a 

 Eke interval, the solar attraction maxima should 

 be "found at l h and 13 h ; the minima, at 7 h and 

 19\ Assuming the attraction tidal curve to 

 be symmetrical, and the deviations from sym- 

 metry as due to differences of temperature 

 and vapor, the author constructs an approx- 

 imate daily barometric tidal table from the 

 Girard College, and a like one from the St. 

 Helena observations ; in each of these are 

 given the 24 hourly values of the rotation, the 

 temperature-and-vapor, and the solar-and-re- 

 sidual tidal elements of the mean daily bar- 

 ometric curve. 



Among the inferences drawn from these 

 tabular results, are: That in the intertropical 

 and middle latitudes the daily temperature tide 

 is smaller than the rotation tide ; that there is 



but one high and one low temperature tide in 

 twenty-four hours ; that the temperature effect 

 upon atmospheric pressure reaches its maximum 

 in the evening, and its minimum in the morn- 

 ing; and that the daily temperature tide in- 

 creases, while the rotation tide diminishes, as 

 we approach the poles. Amer. Jour, of Sci- 

 ence, Nov., 1864. 



A New Water-Barometer. This instrument, 

 constructed by Mr. Alfred Bird, of Birmingham, 

 and which has been in perfect operation for six 

 years, was exhibited before the British Associ- 

 ation in 1865. From the detailed account of 

 its construction and the mode of filling, later 

 given (Philos. Magas., NOT., 1865), an outline 

 only can here be presented. 



The author lays down four things as requir- 

 ing attention in the construction of a water- 

 barometer : 1, that the water must be deprived 

 of air ; 2, that the air must not again enter the 

 water; 3, that the water must go into the ba- 

 rometer tube to the exclusion of air ; 4, that, 

 while the atmospheric pressure must be allowed 

 to act freely upon the water-column in the 

 tube, no air must penetrate into the vacuum- 

 chamber. 



A half-inch white-metal tube was secured in 

 an upright position within a stairway, including 

 near its uppermost part, where the range of the 

 fluctuations of the water-surface must occur, 

 six feet in length of a one-inch glass tube, suit- 

 ably jointed into the former by sockets at its 

 ends ; while alongside the glass tube a scale was 

 fixed, graduated to inches and tenths of an inch. 

 From the top of the scale, a vertical height of 

 422 inches (35 feet) was measured downward, 

 to a zero point. The metal tube above the 

 glass, after passing coiled as a worm within 

 an upright vessel, rises still higher, and being 

 furnished near its top with a closing tap (say, 

 A), it is bent over and terminates open. The 

 lower metal tube, making a curve below the 

 zero point, rises slightly again, and is then bent 

 down so as to -pass into the neck of a one- 

 gallon glass bottle serving as a cistern, and so 

 as to open near the bottom of the latter : at a 

 point just without the neck of the cistern, the 

 tube is furnished with a second tap (B). Into 

 the lowest bend of the tube is inserted a small 

 and short upright pipe, having in its course a 

 third tap (0). On one of the upright guides 

 between which the glass cistern can be raised 

 or lowered by a set screw working from be- 

 neath, is marked the zero point (liquid-surface 

 level for the cistern) of the instrument. 



Four gallons of water having been carefully 

 distilled, this is then in a clean can and beneath 

 two quarts of olive oil, boiled for an hour, to 

 expel the last remains of air ; the can is there- 

 upon sealed, and the contents allowed to cool. 

 The can is then placed at the top of the barom- 

 eter : a long gutta-percha tube, inserted at one 

 end through the oil to near the bottom of 

 the water in the can, is exhausted of air, so 

 that it fills with water to its lower end ; a fourth 

 tap (Z>) near to this is then closed, and connec- 



