12 



NATURE 



[Nov. I, 1877 



latitudes. The opportunities of observing were not 

 frequent, either from the rarity of the phenomena (which 

 is very possible) or because the dense mass of cloud which 

 is the prevailing feature of those regions prevented their 

 being seen except when exceptionally bright. 



Altogether four appearances were noted. The first was 

 1.30 on the morning of February 9, 1874, in lat. 57° S. 

 and long. 75° E., bar. 29-0 in., ther. 35°. There were 

 brilliant streaks to the westward ; no spectroscopic obser- 

 vations were taken. The second was on February 21 at 

 9.30 P.M., lat. 64° S., long. 89" E., bar. 28 8, ther. 31° ; one 

 bright white curved streamer extended from Jupiter, 

 which appeared to be near the focus, through Orion and 

 about as far beyond. Under this was what appeared to 

 be a black cloud, but the stars were visible through it. 

 Real cumulus clouds hid great part of the remainder, but 

 there were two vertical flashing rays that moved slowly to 

 the right (west), generally the aurora was still and bright. 

 On examining the streamer with the spectroscope I 

 found the usual three prominent lines, namely, one 

 yellow- green, one green, the third blue or purple. I 

 looked for the red line but could not find it. 



The third aurora was seen on March 3, lat. 53° 30' 

 S., long. 109° E., bar. 29-1, ther. 36°, after some days 

 wet and stormy weather. Soon after 8 P.M. the sky 

 began to clear and the moon shone out. Noticing the 

 light to the southward to be particularly bright I applied 

 the spectroscope and found the distinguishing auroral 

 line. About midnight I was called as there were very 

 brilliant auroral clouds. The sky was almost clear, but 

 so'.ith were two or three brilliant light clouds, colour very 

 white yellow, shape cumulus stratus ; from about west to 

 near south extended a long feathery light of the same 

 colour, parallel with the horizon, and between south and 

 west there appeared occasionally briUiant small clouds, 

 the upper edges seemed hairy, and gave one the idea of 

 a bright light behind a cloud. The forms changed, but I 

 did not notice any particular order, perhaps because my 

 attention was particularly directed to examining the light 

 with the spectroscope, and the great cold, for my fingers 

 seemed almost frozen, and the motion of the ship made 

 my task rather difficult. I could trace four lines, three 

 bright, and one rather faint, and by reference to the moon, 

 which was shining brightly, roughly determined their 

 places. They must have been exceedingly bright to show 

 so plainly in full moon. The spectroscope used was one 

 of Grubb's single prism with long collimator, A needle 

 point in the eye-piece marked the position of the lines, 

 and a conesponding needle point carried on a frame with 

 the point in the eye-piece and moved by a coarse thread 

 screw, scratched the lines on a plate of blackened glass. 

 I took two plates ; — on the first I scratched the auroral 

 lines and the telluric lines visible in the moonlight ; on 

 the second I scratched the auroral lines, the telluric lines 

 shown by the moon, and the lines given by carbon in the 

 flame of a spirit lamp ; the next morning I verified the 

 hnes in sunUght. The lines marked A are those shown 



A A A k 



\D 



■Car. 



•:F 



■G 



Car. 



Car. 



A<^ 



A 



by the aurora, those marked D, b, F, and G are the 

 telluric lines, and those marked car. were given by the 

 carbon in the spirit lamp. 



The spectrum has been magnified five times from the 

 plates. I cannot account for the different position of the 



auroral lines in the two plates, as the prism was not 

 moved during the observations that I am aware of. 



The fourth aurora was a slight one seen to the south- 

 ward on March 6 at 8 P.M. It would be worth investi- 

 gating whether the low barometer has anything to do 

 with the absence of red in the spectrum, the normal state 

 of the barometer is an inch lower in those regions than in 

 more temperate latitudes. 



I may as well add that on February 9 the aurora was 

 preceded by a watery sunset, and the day broke after- 

 wards with high cirrus clouds and clear horizon. On 

 February 21 the aurora preceded a fine morning, cumulus 

 stratus clouds. On March 3 there was a brilliant sunset 

 followed by a fine morning ; and on March 6, after the 

 slight appearance of aurora, the clouds chansjed to high 

 cirrus. J- P- Maclear 



ABSOLUTE PITCH 



AT the present time the question of absolute pitch is 

 attracting attention in consequence of the discrepancy 

 between Konig's scale and the numbers determined by 

 Appunn's tonometer. This instrument is founded upon the 

 same idea as Scheibler's fork tonometer, and consists of a 

 s.ries of sixty-five harmonium reeds, bridging over an 

 entire octave, and so tuned that each reed gives with its 

 immediate neighbours four beats per second. The appli- 

 cation to determine absolute pitch, however, does not 

 require precision of tuning, all that is necessaiy being to 

 count with sufficient accuracy the number of beats per 

 second between each pair of consecutive reeds. The sum 

 of all these numbers gives the difference of frequencies of 

 vibration between the first reed and its octave, which is, 

 of course, the same as the frequency of the first reed 

 itself. 



The whole question of musical pitch has recently been 

 discussed with great care by Mr. Ellis, in a paper read 

 before the Society of Arts (May 23, 1877). He finds by 

 original observation with Appunn's instrument 258*4 as 

 the actual frequency of a Konig's 256 fork, and Prof. 

 Preyer, of Jena, has arrived at a similar result (258 2). 

 On the other hand. Prof. Mayer in America, and Prof. 

 Macleod in this country, using other methods, have 

 obtained numbers not differing materially from Konig's. 

 The discrepancy is so considerable that it cannot well be 

 attributed to casual errors of experiment ; it seems rather 

 to point to some defect in principle in the method 

 employed. Now it appears to me that there is such a 

 theoretical defect in the reed tonometer, arising from a 

 sensible mutual action of the reeds. The use of the 

 instrument to determine absolute frequencies assumes 

 that the pitch of each reed is the same, whether it be 

 sounding with the reed above, or with the reed below ; 

 and the results arrived at would oe vitiated by any mutual 

 influence. In consequence of the ill- understood opera- 

 tion of the wind, it is difficult to predict the character of 

 the mutual influence with certainty ; but (" Theory of 

 Sound," §§ 112-115) there is reason to think that the 

 sounds would repel one another, so that the frequency of 

 the beats heard when both reeds are sounding, exceeds the 

 difference of the frequencies of the reeds when sounding 

 singly. However this may be, in view of the proximity 

 of consecutive reeds and of the near approach to unison,^ 

 the assumption of complete independence could only be 

 justified by actual observation, and this would be a matter 

 of some delicacy. If the mutual influence be uniform 

 over the octave it would require a difference of one beat 

 per minute only to reconcile Konig's and Appunn's 

 numbers. 



As to the amount of the influence I am not in a position 

 to speak with confidence, but I may mention an obser- 



I It must not be forgotten that the vibration of the tongue involves a 

 transference of the centre of inertia, so that there is a direct tendency to 

 sst the sounding-board into motion. 



