METEORS OF AUGUST AND NOVEMBER. JQ^ 



a similar -uniformity of the other component, namely, their true velocity and di- 

 rection, which, with that of the observer taken with a contrary sign, gives their 

 relative velocity and direction as a resultant; for either there must be this uni- 

 formity, in their true velocities, or a compensation between the true velocities 

 and directions — such as will produce a uniform true component in length and 

 position. Now, the latter circumstance is extremely improbable. Its occasional 

 occurrence might be considered accidental; but its repetition in nine hundred out 

 of one thousand instances in one night, (the exceptions amounting to the tenth 

 or twentieth part being readily accounted for,) indicates the prevalence of a 

 general law, and we are thence compelled to suppose the true velocities and 

 directions the same at the successive anniversaries. If this be admitted, then 

 these successive groups, seen at yearly intervals, are moving, in each instance, 

 in the same part of the same orhit. This identity of orbits is thus established. 

 It will be seen in the sequel that there are only three independent variable 

 elliptic elements of a meteor seen in proximity with the observer. These are 

 the inclination i, the mean daily motion n, and the angle of eccentricity ^; the 

 three remaining elements (the ascending node Si, perihelion n, and epoch H,) 

 being known functions of the first three, and of the meteors' given position in 

 the system. Now we have already arrived, by induction, at the conclusion 

 that the true velocity g is common in quantity and direction. An induction 

 precisely similar, but having less force on account of the greater number of 

 variables, and consequently of possible combinations, compells us to admit 

 that these three elements, i, n, and ^, are common to the successive convergent 

 meteors. For though the same value of g in quantity and direction might 

 sometimes recur on the principle of a compensation of the discrepancies of i, 

 n, and ^, yet the uniformity of such recurrence (with only such exceptions as 

 are referable to another law known to prevail) points to a uniformity of cause 

 to be found only in the identity (or perfect similarity) of the successive com- 

 binations of z, n, and ^, that is of the orbits of the successive individuals of the 

 meteor group. Again, suppose, as in Table II., this principle of similarity of 

 orbits established in this way for the individuals composing the respective 

 flocks seen at Berlin and Breslaw in 1837, and also for those seen at Berlin 

 and Konigsberg in 1839. Then, since the three dependent elements, SI, rr, 

 and jff, have no new feature in 1839, the other three elements, ?, n, and ^, must 

 be supposed to be common to the two phenomena of 1837 and 1839, or we 



