NOVEMBEE 13, 1896.] 



SCIENCE. 



719 



ocean results in enlargement of the lower 

 end of the funnel by the formation of an ir- 

 regularly shaped cloud of vapor or spray. 

 After the disappearance of this waterspout 

 two others, both less perfectly developed, 

 appeared in succession. These were also 

 photographed, but the views are not so 

 striking as in the first case. No authentic 

 scientific account of these waterspouts has, 

 so far as we have seen, been published. 

 The 8 a. m. weather map of August 19th, 

 shows an area of high pressure central 

 north of the Lakes, with cool north and 

 northwest winds over New England. It 

 seems likely, therefore, that these water- 

 spouts were due in part to the low temper- 

 atures brought by the winds. Most of the 

 waterspouts which occur over the Gulf 

 Stream in winter are produced under simi- 

 lar conditions of cool ofishore northwest 

 winds, although, of course, the temperature 

 contrasts are much more marked in winter 

 than in summer. 



SEVEN- DAY THUNDEKSTOKM PERIODICITY. 



Kassner, of Berlin, has been investiga- 

 ting the matter of thunderstorm periodicity 

 in Germany during the past few years and 

 has reached some interesting results. In 

 1893 (Das Wetter, 1893, 12-16) he found 

 that for the period 1883-92 the thunder- 

 storms of Berlin showed a maximum fre- 

 quency on Thursdays and a minimum on 

 Mondays. Further study of the records for 

 Berlin from 1830-40 and 1848-92 indicated 

 a Saturday maximum and a Sunday mini- 

 mum. Polls found that the thunder- 

 storms of Aix-la-Chappelle also had a 

 Saturday maximum and a Sunday mini- 

 mum. This fact was naturally supposed 

 to be connected in some way with the 

 increased smoke resulting from extensive 

 firing up in factories and foundries on Sat- 

 urday, when much work often still remains 

 to be done and has to be hurried through, 

 or, as in the case of iron foundries, the 



metal is melted on that day in order that it 

 may cool over Sunday. On Sunday, on the 

 other hand, there is little smoke because 

 little use is then made of fires. In Das 

 Wetter for August and September, 1896, 

 Kassner has continued his inquiry, using 

 data for other places, and reaches the con- 

 clusion that in general the frequency of 

 thunderstorms increases everywhere from 

 Monday to Tuesday, and that a minimum 

 occurs on Thursday or the next adjoining 

 days. In cities with extended industries 

 which require fires on a large scale there is 

 everywhere an increase from Friday to 

 Saturday, and a decrease from Saturday to 

 Sunday, while in places without many 

 factories the reverse is usually the case. 

 The variations in atmospheric electricity are 

 believed to stand in close relation to the 

 variations in the amount of smoke, as pre- 

 viously suggested by Arrhenius and Ekholm 

 (1894). E. DeC. Ward. 



Haevard University. 



NOTES ON INORGANIC CEEMISTBY. 



The manufacture of acetylene, at a low 

 price, from calcium carbid, and its remark- 

 able power as an illuminant, promised rapid 

 development of its production upon a com- 

 mercial scale. Two schemes are in use for 

 its distribution. In one the plant is local, 

 acetylene being made on the premises 

 where it is used. In the other the acety- 

 lene is made on a larger scale and delivered 

 in steel cylinders, condensed under high 

 pressure. 



The fact that acetylene is an endother- 

 mic compound has given rise to fear that 

 grave danger attends its use, especially 

 when stored in cylinders under high pres- 

 sure. Hence recent experiments of Berthe- 

 lot and Vieille (Comptes Een'dus, CXXIII. : 

 523) on the explosive properties of acety- 

 lene possess considerable interest. They 

 find that, at atmospheric pressure, a decom- 

 position induced by ignition, or by the ex- 



