March 30, 1893] 



NATURE 



511 



stain from any generalisations, it is obvious that the notion of 

 the rotatory power of saline solutions being independent of the 

 particular metal present in the salt is altogether untenable. 



Percy F. Frankland. 

 University College, Dundee, March ii. 



The notice referred to by Prof. Percy F. Frankland was 

 written, and the proof returned to the printer, before the end 

 of last year. Since then two researches have been published 

 — by Purdie and Walker, and by Frankland and Appleyard— 

 in which facts are adduced, apparently irreconcilable with 

 Guye's theory. Had these facts been at my disposal I should 

 doubtless have expressed myself more guardedly. 



Prof. Frankland says : "As far as I am aware, there is npt 

 a single instance of an asymmetric carbon atom attached to 

 four groups qualitatively distinct, being found optically inactive 

 in consequence of two of those groups being quantitatively 

 equal in viass ; " and he complains that I have hardly empha- 

 sised this sufficiently. My reason was, that I was not altogether 

 convinced of the fact, as may be seen from the following 

 passage, which I transcribe, and which originally formed a 

 footnote to the notice in question : — 



"The present reviewer ventures to suggest that cases such 

 as are sought by Guyc are to be found in those compounds in 

 which two of the four different groups attached to an asymmetric 

 carbon atom are themselves asymmetric carbon atoms of equal 

 and opposite enantiomorphism. Such compounds would exist 

 in two distinct forms ; but as the two opposite enantiomorphic 

 groups would be of equal mass and would be situated at equal 

 distances from the central asymmetric carbon atom to which 

 they are attached (inasmuch as the two opposite enantiomorphic 

 modifications of a compound always have the same molecular 

 volume), the conditions necessary for optical inactivity according 

 to Guye's theory would be fulfilled, and neither of the two forms 

 ought to cause rotation of the polarised ray. Such a case has 

 already been observed in the two inactive, non-racemic trihy- 

 droxyglutaric acids described by Emil Fischer (Bet: der deutsch. 

 chein. Ges. 24, p. 4214), although it does not appear to have 

 been hitherto interpreted from this point of view." 



I afterwards suppressed this footnote, partly because it seemed 

 to me out of place in such a notice, and partly because the 

 optical activity of the two trihydroxyglutaric acids could be 

 accounted for in another way : namely, by the fact that, as 

 pointed out by E. Fischer, the mirror images of their molecules 

 are congruent with the molecules themselves. But the passage 

 will show why I was indisposed to enter a proved negative 

 against Guye's theory. 



As regards the charge of "endorsing special pleading " in 

 the interests of the electrolytic theory of Arrhenius by sup- 

 pressing the fact that tartar emetic has, in solution, a different 

 rotation from the other metallic tartrates, I may say at once 

 that I was ignorant of this fact. I am not a specialist on the 

 subject of the optical properties of organic compounds, and I 

 merely summarised, doubtless uncritically, the account of Oude- 

 mans' law given in van't Hoff's book. Indeed, the brief 

 notice, as its wording everywhere indicated, was a summary 

 rather than a criticism. 



I take this opportunity of rectifying an omission. At the 

 time of writing the notice I was not aware that Prof. Crum 

 Brown had, independently of M. Guye, put forward, in the 

 Proceedings of the Royal Society of Edinburgh, views on the 

 influence of the various substituting radicles in modifying the 

 optical rotation of organic compounds. F. R. Japp. 



University of Aberdeen, March 18. 



Standard Barometry. 

 The question of absolute accuracy in barometer readings is one 

 of great importance to meteorologists ; but there has been so 

 much uncertainty shown by the accumulated facts relating to the 

 subject, that I think that no one who has carefully studied the 

 matter has felt fully satl-^fied that strictly comparable inter- 

 national standards had been obtained. An uncertainty of at 

 least OT mm. was indicated by the various international com- 

 parisons of normal barometers which have been carefully made 

 and discussed during the past ten years. I think that at last a 

 definite conclusion has been reached, and that the very recent 

 results published in paper No. 4, Band xvi. of the Reper- 

 ^orium///r Meteoroloifie will he aiccepted AS proving that at St. 

 Petersburg at least normal readings are obtained. 



NO. 1222, VOL. 47] 



About twenty-five years ago Director Wild, of the Central 

 Physical Obi-ervatory at St. Petersburg, established the first 

 normal barometer of the modern form ; and as much as twenty 

 years ago he claimed to have obtained practically normal read- 

 ings. Moreover, he urged that the transfer of these normal 

 readings from place to place by means of portable barometers 

 was impossible within the desired limits of accuracy, and that 

 each country ought to hrve its own thoroughly investigated 

 normal barometer. This last has been proved by the results 

 obtained by various investigators ; and now Prof. Wild offers 

 the proof of the accuracy of his normal barometer in the paper 

 just referred to, which bears the title "Die normal-barometer 

 des Physikalischen Central- Observatoriums zu St. Petersburg." 



This paper was presented to the Academy of Sciences on 

 November 4, 1892, and in it Wild gives the results of the inter- 

 comparison of three local normal barometers. 



Normal barometer No. I. was mounted at St, Petersburg in 

 1870, and was fully described in Band iii. of the Repert. f. 

 Meteor. 



A second normal barometer was mounted at Pawlowsk (about 

 twenty miles from St. Petersburg) in 1887, and a third normal 

 was mounted at St. Petersburg in 1891, and is known as normal 

 No. II. 



In 1887 and 1888 Wild found that the St. Petersburg normal 

 I. and the Pawlowsk normal did not differ by more than o 01 

 mm. 



In 1892 the St. Petersburg normals I. and II. were found 

 to agree within the limit of error of observation (less than 

 O'Oi mm.). 



In 1892 the St. Petersburg Normal II, was dismounted, 

 taken to Pawlowsk, and there compared with the Pawlowsk 

 normal, and the two were found to differ by only O'OI or 0*02 

 mm. ; that is '004 or '008 inch. It must be added that these 

 comparisons have all been checked by means of comparisons 

 with portable barometers of the highest class. 



The paper by Prof. Wild is accompanied by illustrations of 

 these various normal barometers. The St. Petersburg normal 

 has recently undergone some alterations, and these are also fully 

 de>cribed. Altogether this is perhaps the most important con- 

 tribution to the subject that has appeared since Prof. Wild's 

 famous memoir of 1873 ; for we can now rest assured that 

 farther refinement is not required by any practical demands. 



It seems to me that now that we are sure of the accuracy of 

 Wild's normal, it is more necessary than ever that we should 

 know with greater certainty its relation to the principal standards 

 of Europe. I desire, therefore, to propose a plan by which a 

 series of comparisons can he carried out for a few places at a 

 very slight expense, and with as much accuracy as portable 

 instruments will permit. In 1883 it became my duty to transport 

 to America, from Hamburg three of the VVild-Fuess portable 

 barometers of the highest grade ; and it was of great importance 

 to take every possible precaution against their being injured or 

 their condition altered in any way so as to affect their readings. 

 I devised a mounting on shipboard which was very satisfactory, 

 and gave me no cause for uneasiness regarding the barometers, 

 even in stormy weather. So many barometers are sent out from 

 England to almost every country that I strongly urge the use of 

 a similar arrangement in all cases where it is desirable to retain 

 an assigned barometer correction. 



The accompanying sketch shows my manner of mounting 

 the barometers. Two small strips of wood, A A, are 

 screwed to the woodwork running lengthwise of the vessel. 

 They are placed about two feet apart, and are inclined at an 

 angle of perhaps 45°. Small leather straps, say 15 inches long, 

 are fastened to these strips by single screws as shown at BB. 

 A rather soft stuffed flat cushion or pillow is now placed against 

 the woodwork (wall) as shown at C. The box containing the 

 barometer is now pressed against the cushion and the two ex- 

 tremities are placed within the grasp of the straps BB. These 

 last are buckled and drawn tight enough to hold the barometer 

 box firmly against the cushion C. The barometer is thus held 

 in such a manner that no ordinary jarring can cause any damage 

 to it, as there is no direct contact with a rigid surface, since the 

 pillow prevents it from touching the wooden strips, and the soft 

 yielding .straps have a spring-like effect. 



The lower part of the sketch shows the barometer box DD in 

 position, with the barometer shown within it. Of course the 

 cistern is held uppermost. On account of the jarring motion of 

 the ship's screw in rough weather, it is desirable to locate the 

 barometers well amidship, and also have the cistern of the baro- 



