454 



NATURE 



[Sept. 23, 1875 



Office referring to the tabulation of temperatures (vol. xii. p. loi). 

 From 1,283 estimations of tenths of seconds, as tabulated by the 

 highly-trained and experienced observers at Greenwich, he shows 

 that the whole seconds estimated were 1 5 per cent, of the whole 

 number, and thereupon remarks that this is precisely the excess 

 of whole seconds that is taken in the review of the work of the 

 Meteorological Office as indisputably proving the carelessness of 

 the tabulations at the Kew Observatory. This is a mistake. 

 Kew was not singled out for criticism because the whole degrees 

 tabulated there amounted to 15 per cent, of the whole number, 

 but because of " the irregularity of the tabulations, more espe- 

 cially as regards the tabulations from day to day." An exami- 

 nation of the tabulations at Kew from day to day shows that 

 there are first-class tabulators in that Observatory, but it also 

 shows there are others whose work is inferior. Thus, in the first 

 published sheet for Kew, viz., January 1874, on seventeen of the 

 days the whole degrees tabulated amounted on each of these 

 days to at least 25 per cent., and the average of the whole seven- 

 teen days reached 31 per cent., or nearly a third of the whole. 

 On the remaining fourteen days of the month the average was 

 14 per cent. Hence the variations of the numbers of whele 

 degrees from month to month, which, as stated in the review, 

 were 172 for January, 87 for February, 127 for March, and 94 

 for April. It is this irregularity in the work of tabulation 

 which has lowered the character of the work done at Kew. 



The averages calculated from 6,696 tabulations showed that 

 the number of whole degrees read off at the seven observatories 

 were 8 4 per cent, of the whole at Stony hurst, I5'0 at Kew, 

 I9"5 at Aberdeen, 21 "2 at Armagh, 237 at Falmouth, 247 at 

 Valencia, and 24*8 at Glasgow. So far as the mere average 

 numbers are concerned, the tabulations at Stonyhurst and Kew 

 are satisfactory ; not so, however, is the work done at the other 

 five observatories, especially the last three, where, on an average 

 of 6,696 tabulations, a fourth part of all the numbers tabulated 

 were whole degrees. For particular months the percentages: are 

 sometimes very large. Thus, at Aberdeen during January 1875, 

 the following are the percentages of the different decimal places 

 of the dry-bulb readings as printed by the Office : — 



From this examination it is seen that 50 per cent, of the whole 

 readings are assigned to two of the decimal places, viz. 'O and "i, 

 of which 39 per cent, are whole degrees. The largest percentages 

 are not, as in the cases adduced by Mr. Plummer, distributed in 

 different parts of the decimal scale, but stand together, viz., 

 •9, 'o, and "I. As regards the column for each particular hour, 

 out of the thirty-one'readings, nineteen whole degrees occur in 

 the 5 A.M. column, eighteen in the 8 p.m. column, sixteen in the 

 5 P.M., fifteen in the 6 a.m., fourteen in the 4 p.m., thirteen in 

 four of the columns, twelve in six, and so on, down to eight 

 whole degrees in one column, and seven in another, than which 

 no fewer whole degrees occur in any column. It is unnecessary 

 to make any remark on these figures. 



The Meteorological Office has published in their Quarterly 

 Weather Reports the monthly extremes of temperature in two 

 forms, viz. in figures, and in curves of temperature. These were 

 compared and the results stated in the review, from which it was 

 shown that as regards the first month's extremes, fourteen in 

 number, there were twelve errors in the numbers as published 

 by the Office ; and as regards the first year's extremes, 168 in 

 all, there occurred forty-one errors of temperature varying from 

 0°'4 to 9° '6, and twenty-two errors as regards the day and nine 

 errors as regards the hour of occurrence. Altogether twenty- 

 nine months have been examined with the general result of an 

 average of fully four errors in stating each month's fourteen 

 extreme temperatures. Now it is on the large proportion of 

 errors made in stating tlie extreme temperatures (for the pre- 

 vention of which one of the twenty-seven regulations for the 

 Director of the Central Observatory was specially designed), 

 taken in connection with such results as those given above foi 

 one of the observatories for January last, that the charge of inac- 

 curacy in this very costly but vitally important part of the work 

 of the Meteorological Office is based. This charge, Mr. Plummer's 

 letter in no way meets. The simple course is to see that this 

 department of the Meteorological Committee's work, including 

 that of the outlying observatories, be brought under some sort of 

 satisfactory control. The Reviewer 



Ocean^Circulation 



As the strength of Mr. CroU's conviction that he has com- 

 pletely demolished the * ' gravitation theory " of oceanic circula- 

 tion by the " crucial test " to which he subjected it before the 

 Geographical Section of the British Association, is not unlikely 

 to influence the minds of some, I shall be glad to be allowed to 

 point out (i) that I have never denied the existence of a horizontal 

 "wind-circulation," and (2) that the doctrine to which he ap- 

 plied his test was not mine, but a creation of his own. For his 

 whole argument was based on the assumption that the ocean 

 is in a state of static equilibrium ; whereas the theory I ad- 

 vocate, which was originally advanced by Lenz, and which Sir 

 William Thomson (in commenting upon Mr. CroU's paper and 

 my reply to it) pronounced to be a matter " not of argument, 

 but of irrefragable demonstration," is, that the ocean never is and 

 never can be in a state of equilibrium, so long as one part of it 

 is subjected to polar cold, and another to equatorial heat ; but 

 that it is in a state of constant endeavour to recover the equili- 

 brium which is as constantly being disturbed. 



If the boiler and water-pipes 01 a heating apparatus be filled 

 with water whose temperature is that of the budding in which ic 

 is placed, the whole mass of fluid is in a state of equilibrium ; 

 but the lighting of the fire beneath the boiler disturbs that equi- 

 librium, and produces a circulation, which will be maintained 

 as long as the water is being alternately heated in the boiler and 

 cooled by the atmosphere of the budding. 



Suppose that the elongated basin of the Mediterranean, instead 

 of lying E. and W., were to be turned N. and S., so that its 

 water, instead of being exposed (as at present) to a practical 

 identity of thermal influences, should be subjected at one end to 

 arctic cold and at the other to almost tropical heat : instead of 

 remaining in its present state of nearly perfect equilibrium, it 

 would have a circulation like that which I have exhibited in the 

 trough-experiment. 



The only objection raised by Mr. CroU which has even a show 

 of validity, is based on the supposed " viscosity " of water, which 

 he asserts to be sufficient to prevent the disturbance of thermal 

 equihbrium from exerting the effect which the "gravitation 

 theory" attributes to it. This assertion has now been com- 

 pletely disproved by the masterly investigations of Mr. Froude ; 

 who has demonstrated experimentally — what the "wave-line 

 theory" of Stokes, Rankine, and Sir William Thomson had 

 rendered probable — that in the resistance to the motion of a ship 

 through the water, the viscosity of the water itselt is so small an 

 element that it may be practically thrown out, water behaving as 

 a nearly "perfect fluid," except where it moves over solid sur- 

 faces. Mr. Froude (in conversation with me) not only sanctioned 

 my conclusion that a constantly renewed disturbance of thermal 

 equilibrium must produce an oceanic circulation, but mentioned 

 as an instance of the very small difference of downward pressure 

 necessary to sustain such a circulation, that he had ascertained 

 by repeated observation at the mouths of harbours, lochs, and 

 fiords, that wherever the water within has its salinity at all 

 reduced by a mixture with fresh water, there is an underflow of 

 sea-water setting inwards, precisely as in the Baltic and^ Black 

 Sea Straits. 



Mr. CroU attempted to draw a further disproof of the ' ' gravi- 

 tation theory " from the Challenger observations on the tempera- 

 ture of the upper strata of the Antarctic Sea, at and near the 

 ice-border. These observations show that a stratum of water of 

 from 32° to 29° overlies a stratum of Jrom 34° to 32° ; which is 

 considered by Mr. CroU as a death-blow to my assumption that 

 the coldest water smks to the bottom. Now, »ince 1 have re- 

 peatedly pointed out that the water of melting field-ice, and 

 d. fortiori that of melting icebergs, wUl float on ordinary sea- 

 water colder than itself, in virtue of its inferior salinity, and since 

 Capt. Nares distinctly speaks of the cold surface-stratum as 

 having this origin, it does seem to me not a little strange that 

 Mr. CroU should have overlooked this consideration. It is 

 obvious that, for the reason just stated, the descent of the cooled 

 surface-stratum cannot take place in the polar summer at or near 

 the margin of the ice: but that it takes place wherever and when- 

 ever the surface-cold is sufficient to check surface-liquefaction, 

 and to cool down wAter of ordinary salinity to a temperature 

 below that of the subjacent stratum, it wiU be hard for Mr. CroU 

 to disprove. 



I cannot but greatly regret that Mr. CroU abstains from 

 subjecting his conclusions on this subject to the test of 

 personal discussion. For if he would bring them (as I have 

 brought my own) under the criticism of the Mathematicians and 

 Physicists of Section A, he would find that, notwithstanding the 



