izing the zinc, while protecting the metal against oxidation. The use of proof assays is necessary, as is a high temperature (1375) which temperature was not available at the time the method fell into disuse. (0) Comparative Studies.-A summary should be made here of the few important studies that have been made by various investigators on comparison of methods for the determination of zinc. As stated in the introduction, most of the comparative studies which have been made are for the reliability of the precipitation volumetric processes for zinc. E. Oliver (71) compared the Galletti ferrocyanide and the Schaffner sodium sulfide precipitation volumetric processes. These processes were also studied by L. L. De Konninck and E. Prost (106). J. H. Hastings (108) discusses Low's and also Waring's modifications of separating zinc for the Galletti ferrocyanide volumetric precipitation process. B. C. Hinman (62) discusses Von Schultz's and Low's methods of determining zinc, and while satisfactory for readily decomposable ores containing little or no iron, he finds that they require modification in ores containing iron, inasmuch as the iron precipitate carries zinc down with it. . C. Wilson (45) compares and discusses the volumetric K,Fe(CN), method with the gravimetric electrolytic method. Nissenson (41) compares the volumetric sodium sulfide and ferrocyanide processes with the gravimetric methods of zinc as sulfide and by electrolysis, obtaining equally good results. Jander and Stuhlman (80) in their study of chemical analysis with membrane filters studied the volumetric methods of Houber, Taylor, Mohr, Pouget, and Mann. Good results

were obtained by dissolving the ZnS precipitate in excess of standard mineral acids, boiling off the HS and titrating with standard alkali hydroxide using methyl orange as an indicator. H,SO, may be used as a standard acid, but 0.2 normal IICI or HBr was found to be somewhat better. In the precipitation of ZnS from solutions prepared from zinc, accurate results were obtained in test analysis when the nearly neutral solution was treated with sodium acetate and then saturated with H,S. Such precipitates could be weighed as ZnS, after heating in hydrogen, but results were high if they were roasted to ZnO. By the use of membrane filters, a precipitate which would require 6 hours to filter and wash, could be handled in 20 minutes.

3. EXPERIMENTAL.

methods of In this critical study on analysis, a sample of zinc nitrate crystals, Zn(NO3)2 6H2O, containing 0.001% Fe, 0.002% Cl, 0.001% SO,, and no lead, was used. The sample was prepared for analysis by dissolving a weighed amount in water made up to certain volume, and aliquote used for each determination. The gravimetric methods for zinc experimented upon were six modifications of the phosphate method, three modifications of the sulfide method, and three modifications of the oxide method. The volumetric methods tried were three modifications of the K ̧Fe (CN), process, using different indicators, and the Percy Walker's volumetric neutralization method. The total number of methods or modifications tried was nineteen. The following table gives the results of this study :

METEORS.

By F. H. LORING.

It is of interest in connection with the nature or origin of meteoric bodies to consider the supposed relationship between certain comets and showers of shooting stars, as they are popularly called, which are swarms of meteors that pass into the outer atmosphere of the earth at an alti tude of about 65 miles. Other matters are also of interest. These bodies are probably very small, as they appear to be completely consumed by the intense heat generated by their friction in passing swiftly into the rarefied atmosphere; but they leave luminous tails, or trains, behind them as they pass. These trains persist for a little time; even for a few minutes, or much longer in some instances (e.g., 40 minutes). Usually, however, the duration is only a few seconds, or even much less to the casual observer. With the aid of a telescope, such durations as just mentioned are seen.

This seems to show that the phenomenon is due to a trail of luminous gas-a kind of phosphorescence in the gaseous atmosphere through which they glide. In fact, a long

tubular luminous cloud appears often to be formed, but it soon vanishes.

A number of spectroscopic observations have been made. It is necessary to distinguish between the spectrum of the train and that of the meteor proper, or the nucleus, as it is termed. A recent summary of this work is given by Miss Mabel Weil (who worked with the late Prof. C. C. Trowbridge), in the Proceedings of the National Academy of Sciences, U.S.A., January, 1924, p. 24 (see abstract in Chemical News, 1924, CXXVIII., p. 122), attention being drawn to the phenomenon of afterglow in gases as being the cause of the luminous meteoric trains; in fact, it is probable that the observed meterological phenomenon is mainly an afterglow effect.

Spectroscopic investigations of the afterglow of nitrogen by Strutt (now the succeeding Lord Rayleigh) and Fowler, point to this similarity, as may be judged from the accompanying table (I.) taken from the article above mentioned. The afterglow of nitrogen had been previously studied by E. P. Lewis in 1900 and C. C. Trowbridge in 1906.

It is of interest here to note the composition of the atmosphere at different heights. C. P. Oliver in his book "Meteors," 1925, p. 125, gives a table showing the distribution of the gases of the atmosphere (exclusive of the small quantities of neon, krypton and xenon), as given by W. J. Humphreys in his "Physics of the Air." This table is here reproduced with the minor change of omitting some of the decimal-place figures, and rounding off some of the values, as it is not assumed that they are very accurate, those above 35 kms. being extrapolated. The horizontal summations are approximately 100.

Table II.

vapour, remain constant as a result of vertical convection; (7) that above 11 kms. where the temperature gradient is slight, and the vertical convection negligible, the several gases are distributed according to their molecular weights. Middle latitudes

are assumed also.

From this table it will be evident that nitrogen at low pressure should be present in appreciable relative quantity at the elevation at which the trains occur, and this agrees in general with experiment.

Meteoric showers appear to radiate from a given point or small area indicated by the constellation in the field of observation, the radial appearance being due to perspective (vanishing point) for the meteors really enter the earth's atmosphere in a common direction, the paths being sensibly parallel to one another in each shower.

Making a general survey, using some figures given by Denning, meteors are seldom seen above 130 miles, but a case of 216 miles appears to be recorded. It has been stated that meteors generally appear at an altitude of 76 miles (122 kms.) and disappear at 51 miles (82 kms.).

Passing now to the comet side of the problem, apart from the nightly occurrence of shooting stars, there are periodic swarms of meteoric bodies which have been found to be connected with certain comets, apparently exclusively, as shown by Table III., taken from Jones' "General Astron0135 omy," 1922, p. 281.

-56

.0081

.0123 0274

.10

10.

13.

8.63

Table III.

.16

41.

१०.

78.

O, 20.99; Ar 0.94; H2 0.01; Ne 0.0015; He 0.00015; CO2 0.03; (2) that water vapour amounts to 1.2 per cent. at the lowest level; and (3) that it decreases at lower temperatures with increase of elevation to a negligibly small amount at or just below 10 kms; (4) that the temperature decreases at the average rate of 6° C. per km. from 11° C. at sea-level to 55° C. at an elevation of 11 kms; (5) that beyond 11 kms above sea level the temperature remains constant at 55° C.; (6) that up to the level of 11 kms. the relative percentages of the several gases, excepting water

The precise connection between these comets and the meteoric showers is better shown from the following tables, as given in Oliver's book cited above. No other comprehensive tables of this particular kind are given. The elements of these tables need not be elucidated here, as the corresponding values are quoted merely to show numerical similarity in support of the argument presented, for it has been stated that there are only four meteoric swarms (here given) connected with comets, about which there can be no adverse contention.

Table IV.

APRIL 20.

236°

18611. 243° 42′

LONGITUDE OF PP.

INCL'N TO PLANE OF 89

PERIHELION DISTANCE 0.955

0

APRIL 18. APRIL 19. APRIL 20. 255° 42′ 248° 54' 240° 34 29 31 30 04 79 46 71 20 77 29 81 29 0.9270 0.8944 0.9270

05

0.8478

03

30 10. PER S 28m PERS. 29m PER S. 30m. PER S

In the above table, p.p. = perihelion point; a.n. ascending node; e = ecliptic; m miles; s = seconds. It will be seen that the agreement between the elements of the comet, and those of the Lyrids on each side is exceedingly good. lowing tables should be noted.

PERIHELION PASSAGE...

Table V.

The fol

NODE....

LONGITUDE OF PERIHELION.

344°

LONGITUDE OF NODE...

138

16

INCLINATION..

115

PERIHELION DISTANCE..

57 0.9643

137 113

27 34 0.9626

Table VI.

LEONIDS.

10th

To complete this part of the article, a paragraph from Heath's "Popular Astronomy," 3rd edition, p. 109, will be presented here: "From many of the known radiants, showers reappear year after year at the same date, showing that the meteors are spread along the entire length of the orbit with some degree of uniformity. Such is the case with the August meteors, the Perseids. Every year, when the earth reaches the point of intersection of its orbit with the Perseid orbit, on August, we encounter a number of these meteors, resulting in a shower of greater or less persistence. A number of Perseids are also to be seen a few days before and after this date owing to the meteors not being collected into a thin stream in the immediate neighbourhood of the linear or theoretical orbit, but spread out for hundreds of miles on each side of it. The Leonids, on the other hand, appear as conspicuous showers only at intervals of several years, though a few members of the swarm are to be seen every year at the date of the shower,14th November. From this it is evident that the Leonid meteors are thickly congregated along a part of their orbit only, the rest of the orbit being just sparsely populated. The crowded portion is probably long enough to occupy two or three years in passing any particular point of the orbit. The shower may, therefore, recur two or three years in succession, after which it will not be conspicuous for thirty-three or thirty-four years, thirty-three and a quarter years being the periodic time of the meteors. Their orbit is a very eccentric one, reaching out beyond the orbit of Uranus."