Pbbruary 4, 1887.] 



SCIEN'CE. 



113 



A plea for civilian control of the U. S. weather- 

 bureau. 



A recent discussion of the value of the signal-ser- 

 Tice -weather-predictions was begun in the Boston 

 Transcript by a letter from a Boston lawyer. A por- 

 tion of the letter is here given : — 



" To the editor of the Transcript. It would seem 

 that it is time to call for a termination of the farce of 

 publishing the official weather-prognostications, at 

 least so far as the neighborhood of Boston is con- 

 cerned. Whoever is in the habit of looking in the 

 morning paper to find what weather is promised for 

 the day must have been much impressed of late with 

 the faculty for getting it all wrong, which the Wash- 

 ington bureau appears to possess. [Here follows a 

 whole list of notable failures within a month.] In 

 •conclusion, I will only ask whether a ' weather-bureau ' 

 which produces such failures as these is worth the 

 cost of its maintenance ? It may claim, indeed, that 

 it has sometimes prophesied right, but a man in a 

 dark closet could not possibly have guessed always 

 "wrong." 



A number of letters followed this from different per- 

 sons, all of which agreed in regard to the inefficiency 

 of the signal-service predictions ; and this, I think, 

 voices the general sentiment of the New England 

 people. I had so frequently heard people last year, 

 when they were speaking of the signal- service pre- 

 dictions, say, ' Anybody could guess at the weather,' 

 that the question presented itself, Why was it, that, in 

 face of the fact that the official bulletins claimed eighty 

 or even ninety per cent of successful verification, 

 the average New Englander had arrived at the con- 

 clusion that the signal service merely guessed at the 

 weather ? It occurred to me that the popular meas- 

 ure of success was not what per cent some arbitrary 

 method of verification gave, but rather how much 

 better were the predictions than those which could 

 be made by people ordinarily without instruments of 

 any kind ? 



In order to test this, I had Frank Brown, an intel- 

 ligent steward of Blue Hill observatory, make weather- 

 predictions at sunset for the following twenty-four 

 hours dn each day from last March to July inclusive. 

 These predictions I recorded when made, and care- 

 fully verified them in accordance with the rules 

 given by the signal service to voluntary observers 

 for verifying the signal-service predictions. I then 

 compared his predictions with those of the signal 

 service, verified in the same manner, and I found 

 that each month he obtained from three to ten per 

 cent higher success than the signal service. 



In order not to confine the test to one person alone, 

 I asked Mr. and Mrs. Davenport, intelligent persons 

 living near Blue Hill, but who claimed to know noth- 

 ing about the science of meteorology, to make 

 weather-predictions during the month of June. 

 These predictions were made at sunset for the twen- 

 ty-four hours beginning at midnight, and were based 

 on the appearance of the sky alone without any in- 

 struments. These predictions were received and 

 recorded when made, and the end of the month 

 showed that the predictions of each, though slightly 

 different, were eighty per cent verified, while the 

 signal-service predictions during the same time were 

 only seventy-seven per cent verified. 



These results clearly show why many people do 

 not regard the signal-service predictions as of 

 value. 



It would occupy too much space to attempt to show 

 why the signal-service method of verification makes 

 them appear to gain such high success : suffice it to 

 say that many of the cases which, according to the 

 rules adopted, must be recorded as successful, are 

 most glaring failures. 



During the last few months I have endeavored to 

 ascertain the causes of the many failures in New 

 England of the signal-service ' indications ; ' and I 

 find in the position of New England between the 

 lakes on one side, and the ocean on the other, I 

 think, a fruitful cause of the failures of the signal 

 service. We find from local observations here in 

 Boston, that, when a storm approaching from the 

 west passes over, the sky begins to clear almost im- 

 mediately after the passage of the line of minimum 

 pressure. But on a synoptic chart it is frequently 

 found, that even though the centre of least pressure 

 is off on the ocean, it is raining or snowing at certain 

 lake stations, such as Marquette, Oswego, etc.; and 

 the explanation is apparent, for the circulation of 

 the wind is such as to drive the air across the great 

 lakes to these stations, where it arrives laden with 

 moisture and ready for precipitation. The signal 

 service, ignoring all local influences, and basing their 

 predictions on the eastward movement of weather- 

 changes, predict over and over again rain or snow 

 for New England, which, under such conditions, sel- 

 dom arrives. 



Again: an area of high pressure, approaching New 

 England from over the Lakes, may be attended by 

 fair weather ; but immediately it arrives over the 

 Gulf Stream, and begins to force air on the land 

 from the north-east or east, rain begins ; and numer- 

 ous failures of the signal service can, I think, be 

 traced to this cause. 



I have not confined my studies of the signal-ser- 

 vice predictions to New England, but have closely 

 watched them over other parts of the country ; and 

 I have become convinced that the predictions are 

 based almost entirely on the eastward movements of 

 weather- changes, with but little regard to local influ- 

 ence, or to the facts elicited by the splendid re- 

 searches during the last ten years of Loomis, Van 

 Bebber, and a host of others. In other words, the 

 science of weather-predicting in the United States 

 has not advanced a stej) since the days of Joseph 

 Henry and Espy. This, I believe, has largely if not 

 entirely resulted from the militarj' control of the 

 weather-bureau. Conventional routine, and action 

 without questioning, is a necessary part of military 

 training, and it has produced its fruits in a blind 

 following of a few rules and a consequent want of 

 advance in military weather-predictions. Not only 

 does the military organization fail to give the best 

 results which might at present be obtained, but I be- 

 lieve it is immensely detrimental to the advance of 

 meteorology to a higher and more scientific position. 

 In Europe the men in charge of the weather-services 

 are scientific men, who not onlj' do their present 

 work well, but, sustained and enthused by their 

 work, are investigating the difficult problems which 

 present themselves, and thus pushing meteorology to 

 a higher and more scientific stand-point. 



Nor do I think the detriment of the military organi- 

 zation ends with the predicting department. I have 

 known personally a number of bright young men, 

 intensely interested, and trained in science and sci- 

 entific methods, who were kept out of the signal 

 service on account of the military organization. 



