286 



NATURE 



[July 21, 1892 



operandi^ Brennand, with his water-motion actinometer, drew 

 up, by an ample series of observations extended over several 

 years, the Table B given in Nature (January 8, 1891, p. 237). 

 The numbers in this table are ratios, and they may be all multi- 

 plied by any number without any real alteration in the table. 

 The unit of chemical action originally started with was the 

 blackness produced by 1 00 grains of a candle burnt at the unit 

 of distance ; and this is the unit which underlies Table B. 

 Brennand early found, as Roscoe found, that the sun has always 

 the same effect at the same altitude in a perfectly clear sky. 

 Hence, in all the later observations the unit was recovered from 

 the sun. 



Thus, to take a series of observations, with the water-motion 

 actinometer, with strips of an unknown (but uniform) paper : 

 first, a strip is placed in the instrument, the sun alone being 

 admitted by the vertical slit, and the sliding shutter is run up ; 

 we thus get a gradually tinted slip beside the gauge marked in 

 seconds. The altitude of the sun is noted ; suppose it 30° ; 

 the number in Table B for this altitude is O'loyo, i.e. the 

 number of seconds which produces unit darkening by sun alone 



at this altitude is seconds = q4 seconds. Then on the 



0-107 

 sun strip a mark is made opposite the 9J seconds graduation on 

 the gauge ; this is the unit blackness for the paper, and any 

 subsequent strip exposed is " read" by marking the point on it 

 which has the same blackness. 



This method of recovering the unit is not sufficient to deter- 

 mine, for instance, whether the sky in England on a certain 

 morning was really clear, i.e. as truly clear as the Dacca cold- 

 weather sky. To determine this particular point, Brennand 

 lately in England burnt 100 grains of a candle (as near as he 

 could get) similar in composition to his Dacca candle, and the 

 result shows' conclusively, by the exact accord of several observa- 

 tions lately made near Taunton with corresponding old observa- 

 tions at Dacca, that in this case the two candles must have 

 produced equal effects. But it is obvious that the candle could 

 only be trusted by these results. The experiments made with 

 the candle were not made to recover the Dacca unit, but to test 

 the candle. The exact agreement in the several results raises 

 the very strongest presumption that the Taunton candle was 

 equal to the Dacca candle. It is, however, possible that the 

 Taunton sky varied for the several observations in exact ratio 

 with a variation in the Taunton candle ; and it can only be said 

 that Brennand's observations on this particular point, so far as 

 they go, support Roscoe's result that the chemical action of the 

 sun is the same at the same altitude in a perfectly clear sky, 

 always and everywhere. 



When the chemical action of the sky (or of some portion of it) 

 on a piece of flat paper is observed, what is measured is the 

 integral of the resolved effects of each sky element. Thus if, as 

 in Roscoe's experiments, the piece of sensitized paper is exposed 

 horizontally, and the effect of the whole sky (the sun being 

 stopped off) is taken, we have the total effect of a ring of the 

 sky distant 9 from the zenith to be multiplied by cos fl, and then 

 the eflfect of all such rings from the zenith to the horizon to be 

 summed. This view of the resultant action suggested to Bren- 

 nand the more original branch of his investigations. He was 

 early led to suspect that the chemical action of the sky varied 

 in different parts of it. He devised an instrument, which he 

 calls the mitrailleuse actinometer, by which he was enabled to 

 prove that the chemical action of the sky is a minimum in the 

 great circle distant 90° from the sun ; for an altitude a of the 

 sun this minimum he calls 4. Brennand then further proves 

 that the chemical action (at the same time) at any other point of 

 the sky distant Q from the sun is then ?„ cosec fl. 



Having established these important laws, Brennand is able, 

 by mathematical process, having 4 given him, to calculate the 

 total effect of any defined portion of the sky on a plane of sensi- 

 tized paper exposed at any given angle. He was thus enabled 

 to compare Roscoe's readings of total diffused daylight on paper 

 exposed horizontally with his own Dacca readings on paper 

 exposed perpendicularly to the direction of the sun. 



These investigations led Brennand to a theoretic value for the 

 duration of twilight, and to the devising a new instrument, the 

 "octant actinometer," by which the fundamental constant ?„ can 

 be observed directly. 



This "octant " actinometer observes one-eighth of the heavens, 

 cut out by three planes at right angles to each other, placed so 

 that the line OS, the intersection of two of the planes, passes 

 through the sun. Owing to the cloudy skies of Taunton, Bren- 



NO. I 186, VOL. 4.6] 



nand (who experiments only with a clear sky) had been able 

 very imperfectly to test this instrument at the time his paper 

 was read to the Royal Society. He has since found that this 

 "octant" may be turned in any way round OS (above the 

 horizon, of course) without altering the reading on either of the 

 three planes of the octant. This " octant," therefore, only re- 

 quires one-fourth of the visible hemisphere round the zenith to be 

 clear, for a good observation. What is more important, it enables 

 the observer, when the sky is clear, and the sun's altitude from 

 30° to 60°, to take an observation of a part of the sky entirely 

 30° from the horizon ; so that the uncertainty arising from haze 

 near the horizon (which could not before be allowed for) may by 

 this capital instrument be avoided, and ?'„ obtained without any 

 integrations or calculations beyond division by a number. 



In the whole of these later developments, Brennand's work is 

 entirely original. Sir Henry Roscoe, following a somewhat 

 different course of inquiry, has made experiments on the 

 chemical action of the sun and sky at different levels above the 

 sea ; and on the total effect during different months or seasons 

 of the English sky with all its cloud, fog, and smoke ; which 

 last is an important practical measure of the climate in its 

 influence on vegetation, and perhaps on human health. 



The researches of Roscoe and Brennand have thus, though 

 overlapping at particular points, extended mainly in different 

 directions. Brennand, in the ground covered by both, puts 

 forward far the more accurate determinations ; his table (B) 

 given in Nature, January 8, 1891, p. 237, professes to be of the 

 same character and value as a table of the constants of 

 refraction, — Brennand has had half a century's experience with 

 the chemistry of photographic paper, and is an excellent 

 mathematician of the old school. Moreover, the three leading 

 actinometric instruments he has devised, the water-motion, the 

 mitrailleuse, and the octant, show him to be possessed of much 

 resource in devising instruments of research. 



INORGANIC SYNTHESIS OF AZOIMIDE, N^H. 

 A METHOD of synthesizing this interesting compound of 

 ■^^ nitrogen and hydrogen, by means of a simple reaction in- 

 volving only purely inorganic substances, has been discovered by 

 Prof. Wislicenus, and is described by him in a communication 

 to the current number (No. 12) of the Berichte of the German 

 Chemical Society. The reactions by which azoimide has hitherto 

 been obtained have all been of an organic nature, and more or 

 less complicated. The mode of preparation described by its 

 original discoverer, Prof. Curtius, in reality depends upon a 

 very simple reaction, that of nitrous acid upon hydrazine, N2H4, 

 the other hydride of nitrogen whose preparation we also owe to 

 Prof. Curtius, 



NHo 

 I " + HNO„ = N3H -F 2H.O. 



NH2 

 Hydrazine, however, has only yet been prepared from its organic 

 derivatives, and moreover it has not been found practicable to 

 actually convert free hydrazine itself by means of nitrous acid 

 into azoimide, only certain organic derivatives being acted upon 

 by nitrous acid with production of azoimide. The perfected 

 mode of preparation described by Prof. Curtius at the close of 

 last year is very briefly as follows. Benzoyl hydrazine, 

 CfiHg. CO. NH. NHo, is first formed by reacting with ethyl 

 benzoate upon hydrazine hydrate : 



CgHs. COOC0H5 -f NaHj.HjO 



" L CfiH,, . CO . NH . NH2 -f CoHgOH + H^O. 



The benzoyl hydrazine is then converted by means of nitrous 

 acid, obtained from a mixture of glacial acetic acid and sodium 

 nitrite, into the benzoyl derivative of azoimide : 



CgHs . CO . NH . NH2 + HNO2 = CgHg . CO . N^ i I -f 2H2O. 



^N 



From benzoyl azoimide the sodium salt of azoimide is next 

 formed by treatment with sodium ethylate : — 



CfiHg . CO . N/ II + CaHgONa 



\n 



= CgHs. COOC2H5 -^ Na - N/ ||. 

 ^N 



