May 2 2, 1879] 



NATURE 



75 



there are three errors ; p. 147, line 7, dele "is r," there 

 are two other errors on this page ; p. 187 we have a 

 vague reference to Boole's Differential Equations, and a 

 misprint lower down ; there are other minor errors easily 

 detected, but when correcting pp. 11410 116, somebody 

 must have had his eyes shut at times or he would not have 

 passed such a number of clerical errors. 



In IV. we have a fresh work, well adapted for the 

 higher forms in schools, though the examples are in some 

 cases difficult. There are good notes, and the whole 

 book may be recommended to students reading for 

 scholarship or for college terminal examinations. We 

 could put our finger upon many a mistake easily detected 

 by an advanced student, so that we should advise junior 

 pupils not to spend too long a time upon the questions if 

 they do not succeed in getting the same answer as is given 

 in the text. In making this statement we are bound to 

 say that the number of mistakes seems to be no greater 

 than is almost inevitable in a first edition. 



The manual V. contains "more than 160 deductions 

 which have been set at public examinations, worked out 

 in full as examples, together with a collection of speci- 

 men examination papers, which have been set at the 

 examinations. Cambridge Mathematical Tripos, London 

 University Matriculation, &c." This fuller title gives a 

 good idea of the scope of the work : it aims at doing for 

 junior students what is done for higher students by 

 McDowell's exercises on Euclid and in Modern Geometry. 

 We have only been able to look into the book casually ; 

 we have found the parts so examined correct and put in 

 such a way that a lad acquainted with the text of Euclid 

 ought to have no difficulty in following the proofs here 

 given. The student has to draw his own figures. The 

 printing is good and so done as to assist the reader in his 

 work. From the initials attached to the preface we 

 should infer that the compiler is Mr. A. T. Fisher, whose 

 " Book of Algebra " in the same series we commended, 

 at the time of its publication, in these columns. 



LETTERS TO THE EDITOR 



[ The Editor does not hold himself responsible for opinions expressed 

 by his correspondents. Neither can he undertake to return, or 

 to correspond with the writers of, rejected manuscripts. No 

 notice is taken of anonymous communications. 



[ The Editor urgently requests correspondents to keep their Utters as 

 short as possible. J he pressure on his space is so great that it 

 is impossible otherwise to ensure the appearance even of com ■ 

 munications containing interesting and novel facts. 1 



Spectrum of Brorsen's Comet 



I HAVE to thank Dr. Marshall Watts for having called atten- 

 tion to a point of some interest with regard to the spectra of 



mets, viz., which of the carbon spectra agrees with the 



iietary spectrum? In the ca=e of Brorsen's comet the most 

 iiii[iortant question wa«, whether the spectrum differed largely 

 from that of other comets, as found by Dr. Huggins in 186S, 

 and not having much leisure at the time of writing to examine 

 the question of the different spectra of carbon, I overlooked the 

 circumstance that the comparison spectrum used by Prof. Young 

 was the first or flame spectrum of carbon. The difference, how- 

 ever, in the positions of the bands in the two spectra of carbon 

 is a quantity which it is not very easy to answer for in the case 

 if a faint cometary spectrum, and it is but a small fraction (less 

 than one-fourth) of the discordance between Dr. Huggins's 

 imeasures in 1868 and those made at the present return of the 

 ;:oniet. 



In the comparisons made at Greenwich the induct ion spark 

 (without Leyden jars) was taken in a vacuum-tube containing 



alcohol vapour at a pressure of I'zmm., and the green comet- 

 band was compared with this spectrum exactly in the manner 

 de.'icribed by Dr. Watts, though practical difficulties of manipu- 

 lation prevented our making comparisons with the flame spectrum, 

 as I wished. In fact the awkward position of the spectroscope 

 in observing the comet below the pole made the observations 

 extremely difficult, and caused great loss of time, so that the 

 results are not so numerous as they would otherwise have been. 

 On April 17 I used a micrometer eye piece, with a movable bar, 

 the breadth of which corresponded to 30 tenth-metres, whilst 

 the slit was of such a width that the line with which the band 

 in the alcohol-spectrum commences was 45 tenth-metres broad. 

 The bar was brought up from the blue end so as just not to hide 

 the less refrangible edge of the comet-band ; the spectrum from 

 the alcohol vacuum-tube was then flashed in, and the less 

 refrangible edge of the carbon-band was found to be just visible 

 beyond the bar. Several comparisons were made in this way, 

 and I estimated that the uncertainty in the determination of the 

 coincidence between the less refrangible edges of the comet and 

 carbon-bands was but a small fraction of the breadth of the bar 

 (30 tenth-metres). I did Hot obtain any micrometer readings. 

 On April 19 and 28 Mr. Maunder, from readings with a bright- 

 line micrometer, found for the position of the bright edge of the 

 comet-band in the green, compared with the centre of the line 

 at the edge of the alcohol-band (wave-length, SI98'3) : — 



Comet-band. Wave-length ^t-j.i. t cv. 



Tenth-metres. inferred ^'^'^ °' S'"- 



April 19 ... o'5 to blue ... 5190 ... 0'009 in. = 16 tenth-metres. 

 28 ... 45 to red ... 5191 ... 0-013 .. =24 „ 



In computing the wave-length of the bright edge of the comet- 

 band, half the breadth of the alcohol-line ( = width of slit) has 

 been applied. In a similar manner the wave-length of the bright 

 edge of the comet-band in the yellow was found to be 2'4 tenth- 

 metres to the red of the edge of the alcohol-band at 56l0'S, or 

 at 5580, allowing for the width of the slit, which was 0-033 in- 

 or 65 tenth-metres. The position of the blue band was esti- 

 mated to be approximately coincident with the blue band of 

 alcohol at 4834, but this determination is very rough indeed. 

 The dispersion used was that of one "half-prism, viz., 20° 

 from A to II, equivalent to four flint prisms of 60° with a mag- 

 nifying power of twelve. In my former letter I, by mistake, 

 gave the dispersion as equivalent to two prisms only, instead of 

 four. The high dispersion used is of course an important ele- 

 ment in estimating the accuracy of the determination, and on 

 comparing afterwards the flame and vacuum-tube spectra of 

 carbon with the width of slit and other conditions of observa- 

 tion the same as on April 17 and 19, I found the two bands so 

 widely separated that it appeared impossible to mistake one for 

 the other in estimating a coincidence. I may add that the 

 spectrum of Coggia's comet also was found to be identical with 

 the second spectrum of carbon.' With regard to Dr. Huggins's 

 observations of Comet II. 1868, and Coggia's comet. Dr. Watts 

 does not give his reasons for the assertion that the comparison 

 spectrum was the first spectrum of carbon. According to the 

 diagram given by Dr. Huggins, the spectrum in olefiant gas is 

 distinctly different from that in olive oil, which I presume is the 

 first spectrum, and the comet-spectnim agrees with the former. 

 As far as I can judge, this is" the spectrum which we have ob- 

 tained in vacuum-tubes, whether they contain alcohol, carbon- 

 oxide, carbon-dioxide, or olefiant gas. I do not wish to enter 

 on the question as to whether the differences in the carbon-spec- 

 tra result from differences of chemical composition or of mole- 

 cular condition depending on temperature, though I may remark 

 that the same vacuum-tube gives quite a different spectrum when 

 Leyden jars are introduced into the circuit. 



W. H. M. Christie 



Royal Observatory, Greenwich, May 17 



End-on Tubes, brought to Bear upon the Carbon and 

 Carbo-Hydrogen Question 



In Nature, vol. xx. p. 28, there is an important paper by 

 Dr. Marshall Watts, touching certain recent observations of 

 carbon spectra so-called, which seems to offer an excellent 

 opportunity for clearing up certain long-disputed points in 

 spectroscopy, and to the satisfaction, I hope, of every one. i 



Firstly, the Doctor alludes to the recent happy case of Prof. 

 Young, of Princeton, U.S., having last month compared the 

 green band of Brorsen's comet w ith the green band of a Bunsen 

 gas burner, and found them identical in spectrum place, thereby 



