A typical feature of the Perseids is to be found in the streaks which frequently mark their course (Fig. 1), and serve an extremely useful purpose in enabling the directions to be registered with great accuracy. The theoretical velocity of these meteors is thirty-eight miles per second, so that they belong to the swiftest class of such bodies, and, as such, would be individually recorded with much difficulty, were it not for the special feature referred to. Their very rapid transient flights would baffle the observer as he stood endeavoring to retain the exact points of beginning and ending; and in the majority of instances he must absolutely fail to get nearer than a mere approximation. Only in cases where the meteors sped from one star to another, or in courses parallel to closely adjoining stars, could the paths be truthfully reproduced on his map. But, fortunately for such investigations, we have no such difficulties to encounter. The phosphorescent line, almost invariably projected on the sky by the nucleus as it rushes along, remains to guide the eye in fixing its position. It is the authentic signature of the meteor gone before, and during the brief span of its endurance the observer knows how to utilize it. It is seldom these streaks last longer than three or four seconds, though in exceptional cases of Perseid fire-balls they have lingered several minutes. The writer found the average 1·8 second from many observations in August, 1880; and the most frequent duration is about two seconds. All the brighter meteors of the shower display them. Mr. Henry Corder, of Writtle, has observed these Perseids with great diligence in recent years, and retained many interesting notes of their peculiarities. Of 910 meteors belonging to this system, which he saw in the years 1871-'79, 526 were accompanied by streaks. These included 158 of the first magnitude, only 15 of which were devoid of streaks; and 243 of the second magnitude, of which 72 were streakless. Among the smaller members the proportion was larger. He found the brightest meteors were generally pale-green, others orange, etc.
The luminous streaks, which are known to be the ordinary characteristic of these shooting-stars, have acquired a special significance from the fact that by their means the radiant point of the shower is capable of being ascertained with remarkable precision. This important element, to be reliably determined, must rest upon a large number of accurately recorded tracts, which intersect (on being prolonged backward) at a well-defined position. Many observers have succeeded in finding this from results of more or less value. Mr. R. P. Greg analyzed all the positions estimated prior to 1876, and gave the average at R. A. 44°, Dec. 56° north; and Major Tupman, from a discussion of his own elaborate observations in the Mediterranean during the years 1869-'71, derived the point 45½° 56°, as the center of 28 sub-radiants. Evidently the two results, being founded on a large number of trustworthy records, and agreeing so closely as they did,
