When the azimuth has not been mentioned, I have assumed that that of the greatest altitude was meant.

The positions of these places are indicated in the following chart (fig. 2) by their initial letters.

2545

45

2630

30

15

35

Mr. Wheeler gives eight or ten seconds for the passage of the meteor over 180 miles of its course. This gives a velocity of about 20m a second.

The Sherburne observation gives six seconds for 115 miles, or 19m a second.

Mr. Hollister gives 7 or 8 seconds for 215 miles, or 29m a second. Mr. Pratt of Binghamton saw the meteor for a course of about 190 miles. He rose from his seat and walked across the street during its flight to avoid losing sight of it behind buildings. Eight or ten weeks later he walked over the same space with what he considered the same speed, while I noted the interval, 15 seconds. 'Had this been done immediately after seeing the meteor it would have afforded by far the best determination of the time of flight. But a person under excitement must move more quickly than he would weeks afterward suppose. interval of 15 seconds gives a velocity of nearly 13 miles.

An

Mr. Crowel who saw it from Sherburne gave 8 seconds as the interval of flight. Mr. Arthur G. Newton who saw it from Parma, N. Y., gave 6 seconds. The same method of obtaining the time was employed in both cases as in the New Haven observation. In the newspaper notices the following periods were assigned: at Roxbury, Mass., 10 or 158; Sharon, Ct., 15 to 20s; New York City, 5s; Perth Amboy, 30 to 45; Brooklyn 10 to 15; Ironton, N. J., 10s to 15; Morris, N. Y., 12s; Riverdale, 30s; Mansfield, 70s; Avoca, 15s; Evensburg, 8 to 10s; Danville, 40; Monticello, 30 to 45s; Rochester, 12.

I give all of the times assigned. But it is evident that the longer periods ought to be rejected. Disregarding those of 30 or more than 30 seconds, the average of the remainder is 11 seconds. Assuming that the observers saw the meteor describe on the average three fourths of its visible trajectory we have a velocity of 16 miles a second. But if we reject also the Sharon and Avoca observations, we have a velocity of 19 miles.

Unless there is a constant error in the method made use of at New Haven and Sherburne for ascertaining the interval of flight, eighteen miles a second must be considered a tolerable approximation to the true geocentric velocity. It is worthy of remark that the observers were in very favorable positions for estimating the time, the motion across the sky being quite uniform.

To obtain the heliocentric velocity it may be well to allow for the effect of the earth's attraction even though the amount of the correction is decidedly less than the probable errors of the observation, and are perhaps less than the effects of the resistance of the atmosphere. The increment of velocity due to the earth's attraction is 1m-4 and the change of direction nearly 5°. These corrections applied to the direction and velocity above given would make the velocity 16m-6 a second, towards R. A. 108° 30', Dec. +43° 45'. The parts of the earth directly under the meteor were moving towards R. A. 42° 7', Dec. +16° 15'. The heliocentric velocity of the meteor is then 30m-4, towards R. A. 67° 45', Dec. +33° 25'.

This of itself indicates a hyperbolic orbit.-Let us therefore inquire whether any changes are admissible that will make the

orbit an ellipse. The path of the meteor cannot be carried much to the eastward so as to shorten its length, and thus diminish the velocity. Its azimuth may be changed, several degrees even, without great violence to the observations, but this will have little effect upon the heliocentric velocity.

The meteor cannot reasonably be supposed to have ascended during the larger part of its course. On the other hand, to change it so as to make it describe a descending line would increase the heliocentric velocity.

Any change then must be in the duration of the flight. If the body was moving about the sun in an elliptic orbit, and the line given above was its true path, the geocentric velocity could not be greater than about 14 miles a second, which would require 16 or 18 seconds for the whole flight. This would moreover make no allowance for the resistance of the atmosphere. Though an elliptic orbit is very possible I can hardly think it probable. Yale College, March, 1862.

ART. XXXIII.-On Orthite from Swampscot, Mass.; by DAVID M. BALCH, B.S.

MORE than a year since, while examining the rocks at Swampscot Beach, near the Clifton House, I observed a mineral occurring in small amorphous masses in quartz and red feldspar, which immediately attracted my attention by its peculiar lustre and appearance; this on examination has proved to be orthite almost identical with that from the Norway granite.

The shore at Swampscot and for half a mile or so N.E. towards Marblehead, consists of ledges of rock, against which the sea breaks, ranging from sienite to porphyry, and extending 50 or 60 feet backward, and, at the most, from 15 to 20 feet in height, to the pasture land above; these rocks are very rugged, seamed with trap, veins of quartz and red feldspar; in the latter occur good specimens of fibrous epidote, and the mineral furnishing the subject for this paper.

The orthite is found almost always imbedded in quartz, but I have obtained a few pieces from the feldspar, which forms the bulk of the veins; it occurs in black amorphous masses, sometimes surrounded with a reddish coating of sesquioxyds of iron and cerium, when much exposed to the weather or the action of the sea; and in very small quantity, for I only obtained a few grammes in return for searching the rocks throughout their whole extent. I examined this mineral shortly after its discovery, sufficiently to determine its name and general composition, and more carefully again this winter, and give below an outline of the method of analysis, and its results.

The orthite of this locality appears always to occur massive, at least no specimens that I have found show any signs of crystal faces. Sp. gr. at 18° C. 3.69-371. Lustre, resinous and in some cases nearly vitreous. Color, jet black, and streak gray. Heated in thin splinters before the blowpipe, fuses slowly to a black blistered glass; dissolves readily in borax and gives a globule, red when hot, but yellow after cooling; with soda gives a slight manganese reaction; in its natural state is very easily decomposed by chlorhydric acid, but after ignition is not affected by it in the least. A small portion being reduced to powder, dried at 110° C., and then ignited, lost 1-49 per cent of combined water.

The preliminary analysis was performed on 445 grms.; from it I ascertained the composition of the mineral, but was unable to determine accurately the amount of those substances present in small quantity. My method of analysis was that given by Wöhler,* in which cerium, yttria, &c., are separated from iron and alumina by moist carbonate of baryta. The following results were obtained:

Magnesia and soda were also present.

32.62

29.47

24.74

8.05

The ceric oxyd, &c., weighing about a decigramme, was exam ined for yttria, and abundant traces of it found; also inconsiderable traces of manganese.

The succeeding analyses undertaken principally to ascertain the amount of yttria present, were conducted as follows: A quantity of the levigated mineral, dried at 110° C., was decomposed in a porcelain dish by chlorhydric acid, (sufficient nitric acid being added during the operation to peroxydize the iron) and the silicic acid separated as usual: the filtrate from the silica was then supersaturated with ammonia by which oxyds of iron, aluminum, cerium and yttrium were thrown down, the magnesia being held in solution by the chlorid of ammonium formed by the excess of chlorhydric acid present in the liquid; this precipitate, which I will designate by a, was completely washed by decantation and then on a filter, and the washings, evaporated to a small bulk, were mixed with the filtrate; from this the lime was thrown down by oxalate of ammonia, and weighed first as carbonate and then as sulphate; in the filtrate from the lime magnesia was precipitated by phosphate of soda, as I did not intend to determine the amount of soda or potassa which may have been present; manganese was also present in such small tracess that it was disregarded.

In order to separate the ceric-oxyd and yttria from sesquioxyd * Practische Uebungen in der chemischen Analyse, Göttingen, 1858, s. 112.

of iron and alumina in precipitate a, I digested it, while still moist in aqueous oxalic acid; the separation was perfect, and the insoluble oxalates obtained were free from iron; after standing 24 hours, these were well washed, dried, ignited and weighed as ceric-oxyd and yttria. The sesquioxyd of iron and the alumina, contained as acid oxalates in the filtrate and washings of the above, were now separated, in the first case as usual, by potassa; in the second analysis by evaporating to dryness, igniting to destroy oxalic acid and expel ammoniacal salts, peroxydizing any reduced iron by nitric acid, and weighing the mixture of AI,O,+Fe,O, as a control to the first analysis.

3

These analyses were performed exactly alike, with the exception given above, the first on 782 grms., and the second on 1.032 grms. of substance.

To separate yttrium and cerium oxyds, I employed the following process given in Rose's Handbuch der analytischen Chemie, Braunschweig, 1851, Bd. 2, s. 72. The ceric-oxyd, &c., from both analyses, weighing nearly five decigrammes, was fused with bisulphate of potassa, the mass digested with a warm solution of sulphate of potassa sufficiently concentrated to deposit crystals on cooling, and the insoluble sulphate of potassa and cerium entirely washed out with this solution; yttría was thrown down from the decanted liquid and filtrate by excess of potassa, and weighed as such; it contained a little manganese.

The following results were obtained by these analyses:

In orthite and allanite about half the iron is usually in the state of sesquioxyd; in analysis A, all the iron is calculated as protoxyd which occasions the apparent loss; an attempt was made to determine the amount of protoxyd present, by the chlorid of gold and sodium process, but it was unsuccessful, and could not be repeated for want of material.

*

These analyses and the physical properties of the mineral, prove it to be orthite, very closely resembling that from Hitteroe, Norway.t

It is interesting to remark that this is the first instance of the occurrence of yttria in New England, if we except the single specimen of yttrocerite, found sometime since in Worcester Co., Dana's Mineralogy, 1854, p. 210.

*Rose's Handbuch, Bd. 11, s. 129.