Mat 14, 1920] 



SCIENCE 



489 



to the northwest, a situatioa that gives rise to cold 

 northerly and northeasterly winds in central Eu- 

 rope. . . . While the inunediiate causes of these 

 interruptions of temperature has thus been made 

 clear, it is not yet certain whether or to what ex- 

 tent such interruptions, with their attendant baro- 

 metric couddtions tend to recur from year to year 

 on certain dates, such as the days of the ice 

 saints. Irregularities in a curve showing the mean 

 annual march of temperature as deduced from a 

 record of 50 or 100 years may be due to excessive 

 departures in particular years rather than to a real 

 tendency to recurrence on particular dates, and, on 

 the other hand, a tendency to recurrence might 

 not manifest itself in the mean curve, especially, if 

 as some students have surmised the phenomenon is 

 one that undergoes periodic fluctuations. 



Bearing on this question is a mathematical 

 discussion by Professor C. F. Marvin, en- 

 titled, "Normal temperatures (daily): are ir- 

 regularities in the annual march of tempera- 

 ture persistent ? "^ Average annual tempera- 

 ture curves based on the averages of the means 

 of each week over a period of years, may be 

 well-covered mathematically in a curve of one 

 or two harmonics. The residuals, which in a 

 given period are much the same over a large 

 part of the eastern United States, are mostly 

 due to some extreme departures occurring in 

 a single year of the record: which throws 

 doubt on the existence of recurrent irregu- 

 larities. 



Professor Marvin's mathematical analysis 

 of only 15-year averages shows that it is pos- 

 sible to get a surprisingly accurate, smoothed, 

 normal annual temperature curve from a 

 short record. 



NOTES 



The Monthly Weather Review^ contains so 

 much material that these occasional notes in 

 Science have by no means covered even a 

 majority of the 150 contributions, not to 

 mention hundreds of abstracts and other items 

 of meteorological interest, published during 

 the past year. Por a brief summary and men- 

 tion of many of the important contributions 

 published during 1919, and the reader is re- 



^ Ibid., pp. 544-555, 4 plates, fig. 

 8 Government Printing OflSce, Washington, D. C, 

 printed for the Weather Bureau. 



f erred to the American Year Book; and for 

 the articles and notes themselves, to the 

 Monthly Weather Review files maintained at 

 all Weather Bureau stations, and at a few 

 hundred college, university and public li- 

 braries. 



Hereafter, these notes on meteorology and 

 climatology for Science will be continued by 

 Mr. C. LeEoy Meisinger, assistant editor of 

 the Monthly Weather Review. 



Chaeles F. Brooks 



Washington, D. 0. 



SPECIAL ARTICLES 



THE SIPHON IN TEXT-BOOKS 



The treatments commonly accorded to the 

 siphon in text-books of physics of college 

 grade may be classified in three groups. I 

 have attempted to reduce the characteristic 

 features of each group to a typical or stand- 



C B L 



ard form. There is no intention to quote 

 and italics are strictly mine. Reference is 

 made to the diagram, which will serve in 

 common for the three methods of treatment. 

 I. The pressure at A is the resultant of an 

 upward pressure equal to the atmospheric 

 pressure and a downward pressure due to the 

 column of liquid AB. The pressure at D is 

 the resultant of an upward pressure equal to 

 the atmospheric pressure and a downward 

 pressure due to the column of liquid DB. As 

 DB is greater than AB, the resultant pressure 



