Sept., 1905] 



KNOWLEDGE & SCIENTIFIC NEWS. 



planation, and it should be thoroutjiily investig-ated be- 

 fore being rejected. 



If it be true that one reaction docs influence another, 

 the importance of the invcstig-ation cannot be over- 

 estimated. The preparations and violent reactions 

 that go on from day to day in a laboratory may be 

 altering (and those accustomed to manage a chemical 

 laboratory know how certain substances do unaccount- 

 ably alter) the molecular arrangement of the substances 

 in the neighbourhood of the demonstation benches. 



This, moreover, raises the question : " Does the 

 weather influence the communication between one 

 chemical reaction and another?" As has been stated 

 above, it was observed that the results of the experi- 

 ments varied (in degree) from day tO' day. 



A thorough investigation of this subject may shed a 

 new and more satisfactory light on the cause of intra- 

 molecular action. 



Among the few experiments described above, there 

 may be some that will suggest others which will lead 

 to more decided and consistent results, so that if it be 

 possible tO' control chemical reactions at a distance, 

 further research would show how it can be most effi- 

 ciently demonstrated. 



A. F. B. 



B. 1. 



Practical Meteorology. 



II.-RainfaLll. 



By William Marriott, F.R.Met.Soc. 



In the present article it is proposed to deal only with 

 the rain after it has reached the earth. In the term 

 " rainfall " is included rain, snow, hail, dew, mist, &:c. 

 The rainfall is always expressed in inches, and is sup- 

 posed to represent the height to which the rain would 

 rise on the level ground if none of the water were per- 

 mitted to run off or percolate through the soil, or to 

 evaporate. 



The instrument used for measuring the rainfall is 

 called a rain gauge. This is best made of copper, and 

 should have a circular funnel of five or eight inches 

 diameter. It is very desirable that it should be of the 

 Snowdon pattern, which has a deep rim tO' retain snow 

 (Fig. i). The gauge should be placed in an open 

 and well-exposed situation free from trees, walls, and 

 buildings; and should be firmly fixed so that it cannot 

 be blown over. The top of the funnel should be one 

 foot above the ground, and be quite level. The 

 measurement of the rain is effected by pouring out the 

 contents of the can or bottle into the glass measure 

 and reading off the division to which the water rises. 

 The gauge must be examined daily. When snow falls, 

 that which is collected in the funnel is to be melted by 

 adding a known quantity of warm water, and entering 

 the difference as rain. 



Rain gauges should not be placed on walls or roofs, 

 as the buildings themselves offer obstructions to the 

 wind which carries the rain drops over the funnel and 

 so gauges mounted in such positions collect less rain 

 than those placed on the ground. This was demon- 

 strated as far back as 1766, for in that year Dr. W. 

 Heberden, F.R.S., had three rain gauges at work at 

 Westminster — one on the roof of the dwarf tower of 

 the Abbey, one on the roof of a house close by, and 



another in the garden of the same house. The amounts 

 of rain collected by these gauges were : — • 



Tower of Westminster Abbey 



Roof of house 



Garden 



1210 inches 

 18-14 ,. 

 2261 



These differences were due almost entirely to the 

 action of the wind. 



Through the influence of Mr. G. J. Symons it was 

 agreed some years ago to adopt 9 a.m. as the hour at 

 which the rainfall should be measured each day, and 

 the amount entered to the previous day. There had 

 been much diversity in this matter, observers measuring 

 the rain at various hours, e.g., 8 a.m., 9 a.m., 10 a.m., 

 noon, 3 p.m., and even midnight. As there are now 

 nearly 4,000 observers in the British Isles, 9 a.m. is 

 evidently the most convenient hour to the vast majority, 

 and its adoption has secured uniformity in the measure- 

 ment of rainfall. 



f 



% 



Fig. I.— Snowdon Pattern Rain Gau?e. 



In hilly and mountainous districts, and in places 

 where it is not possible to visit the rain gauge daily, 

 the contents of the gauges should be measured monthly, 

 the morning of the ist of the following month being 

 chosen for the purpose. These mountain gauges must 

 be of sufficiently large capacity to contain the month's 

 rainfall. 



As everyone knows, the rainfall is very irregular, 

 but, as a rule, there is most rain in the autumn and 

 winter, and least in the spring. The following figures 

 gi\e the average monthly rainfall at the Royal Ob- 

 servatory, Greenwich, for the 89 years, 181 5-1903 : — 



It will thus be seen (Fig. 2) that October is the 

 wettest month with 2.72 ins., and that February and 

 March are the driest months with 1.52 ins. each. 

 Although the above values represent the average rain- 

 fall, the individual monthly falls are often greatly differ- 

 ent. For instance, with regard to the month of 

 October, the fall in 1834 was only 0.47 in., whilst in 

 1880 the fall was as much as 7.65 ins. Again, with 



