January 23, 1891.] 



SCIENCE. 



47 



would seem to be indisputable, then, that the American people 

 are satisfied with their buildings if the outsides are good-looking. 

 The structures illustrated in the Record and Guide include private 

 residences, apartment-houses, hotels, warehouses, and churches, 

 any one of which must have required some ingenuity in arrange- 

 ment of plan, and have had some interesting constructive details, 

 but they are carefully hidden from those who should be interested 

 in these essential portions of architecture. 



These indications of the tendency of American architecture 

 show very clearly where the error is. The needs of the public are 

 heeded in almost every phase of modern life and thought. The 

 manufacturer and the shop-keeper, not less than the editor and 

 the artist, are continually on the lookout for what the public 

 wants, and hasten to supply them as soon as manifested. The 

 public evidently want only exteriors in architecture. Plans, use, 

 environment, and other matters which were once pre-eminent in 

 the art, are now at a discount. Until the popular mind frees itself 

 from such erroneous ideas, it will be impossible for the art to make 

 any progress. It is well to remember that the general public which 

 is satisfied with such things is more to blame for their continuance 

 than the architects who prepare the designs ; but it is a serious 

 retrogression when the architects join the popular movement, and 

 give their assent and support to it by catering to its most objec- 

 tionable features. Baee Ferree. 



School ot ArcUitecture, University of Pennsylvania, Jan. 8. 



Cyclones and Anticyclones. 



It seems to me that the discussion in regard to the origin of 

 cyclones and anticyclones that has been in progress in Science and 

 other journals for several months past opens up a question that 

 has so long been regarded as settled, that it seems impossible to look 

 upon it as being in doubt. It is, in short, as to whether gravitation 

 is the chief cause of movements of the air. Barometric observa- 

 tions have directed attention so forcibly to the relative weights of 

 .columns of air in storm-centres and elsewhere, that it has been 

 assumed as a matter of course that the pressure gradients thus 

 made manifest ai'e the occasion of the horizontal movement ap- 

 parent as wind. If this be the true explanation, in order that 

 such horizontal movement may continue, it is necessary that there 

 be a corresponding vertical movement, and that it be sustained by 

 adequate renewal of the buoyancy of the air in the proper locali- 

 ties. This renewal of buoyancy can only be accomplished, so far 

 as our knowledge at present extends, by heating. But now we 

 are informed as a matter of fact that the air at anticyclonic 

 centres descends in spite of its being warmer at an elevation, and 

 in like manner above cyclonic centres fails to descend, although 

 colder than at the surface of the earth. This certainly opens up 

 the entire question as to whether there is ascensional movement 

 at storm-centres commensurate with the extent and velocity of 

 the winds blowing horizontally, and supposed to be due to an in- 

 draught; or, in other words, whether gravitation really plays the 

 part that has been tacitly assigned to it, or whether it must be 

 relegated to a subordinate position. Personally I am very glad 

 indeed that a discussion having such bearings has come up at this 

 particular juncture, because it has increased very decidedly my 

 interest in following certain clews that look promising in regard 

 to the effects of variations of the earth's magnetic condition as a 

 whole. M. A. Veedee. 



Lyons, N.Y., Jan. 5. 



Dr. Hann and the Condensation Theory of Storms. 

 The time has not yet come for a review of the various discus- 

 sions upon this subject that have been published during the past 

 four years. I doubt if there has ever been a better illustration, in 

 the history of meteorology, of the absolute necessity there exists 

 of appealing to observations in order to establish intricate theories, 

 than the recent discussions on the reversal of temperature in our 

 storms and "highs," which is but another way of putting the 

 problem before us. In this very line Professor Davis says {Science, 

 Jan. 3), "Records of temperature made on high mountain-peaks 

 furnish the best means of testing the convectional theory of 

 cyclones ; for, even if all other tests were successfully borne, failure 



under this test would be fatal to the theory." This statement of 

 the case should be received with a little caution, however, because 

 the presence of the mountain must be a modifying cause, and 

 oftentimes there are cases in which some part of the storm, or 

 high, has its action below the mountain-peaks (I have found this 

 true especially at Pike's Peak) ; but the larger commotions of the 

 atmosphere may be profitably studied at such points. 



In carrying out my studies on this problem, I have invariably 

 sought for help from the original records, which are now so abun- 

 dant at Mount Washington, Pike's Peak, and at many high sta- 

 tions in Europe, and T have massed thousands of observations 

 bearing on the question. The first publication of these studies 

 was in the American Meteorological Journal of August, 1886, in 

 which I showed that the temperature observations at the base and 

 summit of Pic du Midi, in France, indicated a decided rise at both 

 points on the approach of a storm. In October of the following 

 year I showed by the observations at Mount Washington that in 

 both storms and highs there was the same fluctuation at the sum- 

 mit as on the base, and that the mean temperature of the air-col- 

 umn was ten to twelve degrees higher in storms, and the same 

 amount lower in highs, than before or after the centre had passed. 



It seems to me that the crucial test in Dr. Hann's recent work, 

 which has attracted so much attention, must be the records at 

 the mountain stations, and I believe that this will be insisted oa 

 by Dr. Hann himself as strongly as by any one. In fact. Dr. 

 Hann has based all his work on his interpretation of the records. 



5 ^3 If n /9 II ;j ^j- ij If 



zo" 



TEMPERATURE FLUCTUATIONS, 1889. 

 Sonnblick, full curve ; Salzburg, dotted curve. 



It seems to me that he has given altogether too much weight to a 

 few isolated cases, while he has ignored hundreds of cases which 

 disprove his propositions. I have already shown in this journal 

 for Sept. 5, 1890, that the evidence at Sonnblick is different only 

 in degree from that in this country, and I have there explained 

 how the peculiar results in the remarkable high of barometer, 1889 

 (which, in fact, was the only one in three years exhibiting such 

 discordances from the usual law), might be accounted for. I 

 have now made a special study of the storm of Oct. 1, 1889, which 

 Dr. Hann advanced as favoring his view, that the temperature in 

 a storm falls as we rise in its centre, and at some height is lower 

 than that of the surrounding region. The results of this investi- 

 gation so remarkably corroborate my position, that I present a 

 copy of the curves in order that others may see the exact state of 

 the case. 



These curves are constructed as follows. The lower or fuU 

 curve represents the temperature observation for each day at 

 Sonnblick, 3,095 metres (10,154 feet), at 9 p.m., at which time very 

 nearly the mean for the twenty-four hours occurs; and the upper 

 or dotted curve shows the temperature at precisely the same time 

 at Salzburg, just north of Sonnblick, at a height of 437 metres (1,434 

 feet). I have given the curves from Sept. 13 to Oct. 5, including 

 the storm of the 1st. It wiU be seen that there is a most remarka- 

 ble accordance between these curves ; almost every bending at the 

 base is faithfully reproduced at the summit; and, if any thing-, 

 there is generally a greater fluctuation on the mountain than on 

 the plain. This is not all, however. Examining the very date 

 under discussion, Oct. 1, we find that at Sonnblick the tempera- 

 ture began rising on Sept. 29, and in twenty-four hours had risea 



