700 EXPERIMENT STATION RECORD. 



"The excess in favor of the forecasts is not so great as I bad at tlrst liopeil, bnt 

 shows, as fnlly as it is possible to do, tliat tlie forecasts have a scientific basis, and 

 are not merely the resnlt of gnesswork." 



Tlie difliculties which have beeu found to impair tlie accuracy of fore- 

 coasting by this method are interference of the Aveather waves, by which 

 one series is destroyed or its phase reversed, and two sets of storms, 

 during October, one coming over New England from the lakes, the 

 other from the South. 



'^ The results were too intricate for my analysis, and produced a series 

 of disheartening failures iu the forecasts. 



"Difficulties are to be expected in a new enterprise, but I am conti- 

 dent that they will be solved, one by one, until au exact science of the 

 weather is constriu-ted." The measure of success already attained is 

 held to be " substantial evidence that the right clue to long-range 

 weather forecasting has beeu found and will some day be perfected." 



The moon and rainfall, H. A. Hazen (Amer. Met. Jour., 11 {1^95)^ 

 Xo. 10^ pp. 37S-375). — Investigations made by the author at New Ilaveu 

 in 1880 showed " that nearly one-half more precipitation occurred at 

 the time of new than of full moon." A study of the records for one 

 hundred years at London yielded negative results, as did that of the 

 rainfall of the whole country. The latter indicated " that the influence, 

 if there were one, must be looked for in a rather circumscribed region." 



The present paper records a study of the rainfall of New York City, 

 New Haven, Connecticut, and IJoston, Massachusetts, arranged accord- 

 ing to each lunation, 290 lunations in 3 groups being represented. 



" While both New York and New Haven show an increase at the time of new moon, 

 yet the figures at Boston show a very remarkable maximum at the day of uew moon, 

 and au equally remarkable minimum on the sevenieenth day of the luu^ition, or two 

 days after full moon. I do not set forth these figures as an absolute proof that the 

 luoou docs infiucnce rainfall in the neighborhood of Boston, but it looks as though 

 there ranst be somctlung in it. There are at times remarkable reversals in the curves 

 whereby a minimum point in one curve coincides with a maxininm in the next, but 

 to offset these there are remarkable coincidences; for example, a marked maximum 

 at the twentieth day of the lunation throughout. It will require a good many more 

 observations to prove anything one way or the other, but it would seem as though 

 the curve at Boston could not l)e very materially moilified Avith less than fifty years' 

 observations." « 



Meteorological observations at Massachusetts Hatch Station, 



C. D. AVarner and V. L. Warren (Massachusetts Hatch fSta. Met. Ihil. 

 72. pp. o). — Daily and monthly summaries of observations during 

 December, 1S!)4, and a general summary for the year. The annual sum- 

 mary is as follows: Pressure (inches). — Actual maxinuim, 30.47, Janu- 

 ary 17, 10 p. m.; actual minimum, 28.51), January 30, G a. m.; mean 

 reduced to sea level, 30.085; annual range, 1.88. Temperature^ (degrees 

 I?.)._Maximum, 97, 98, ,Iuly 20; minimum, —9,-17, February 25 ; mean, 

 49.7, 47.9; annualrange, 106, 115; maximum mean daily, 82.5, 84, July 



'The first figures denote reading at top of tower (51 ft. above the ground), the 

 second at 'uase. 



