20 EXPERIMENT STATION RECORD. 



follows: For Lake Superior, 35 bottles recovered in 1892 ami 30 in 

 l.*^03; Lake Michigan, 1G3 in 1892 and 35 in 1893; Lake Huron, 142 in 

 1892 and 48 in 1893; Lake Eric, CO in 1892 and 30 in 1893; and Lake 

 Ontario, 55 in 1892 and 25 in 1893. 



Protection from lightning, A. McAdie {U. S. Dept. Agr., Weather 

 Jiiirean Circular, April, ISHi, pp. 20, pis. 11). — This is a circular of 

 information, giving "statistics of actual losses, tbe theory of protec- 

 tion in language free from technicalities, a collection of practical rules 

 for guidance in selecting and maintaining conductors, and, finally, a 

 notable illustration of the successful use of protectors," viz, the Wasli- 

 ington Monument at Washington, D. C. 



The statistics reported show that in the United States during the 4 

 years 1890-93, 784 deaths due to lightning occurred — practically all 

 during the 6 months April to September — the maximum deatb rate 

 being in June and July. During the 8 years ending with 1892, 3,510 

 tires due to lightning occurred, involving a loss of $12,003,835. Dur- 

 ing 9 years ending 1892, 2,335 barns, 104 churches, and 004 dwellings 

 have been struck by lightning. 



"One inteTesting point which appears to be shown by statistical studies of light- 

 ning stroke is tlie decreased liability to accident in thickly settled coninmnities. 



" It may be said, in general, that the risk in the country is live times greater than in 

 the city. For ordinary dwelling houses, not unduly exposed in city blocks, lightning 

 rods are hardly necessary, a very considerable protection being aftorded by the tin 

 roofing, numerous' cornices, gutters, etc. The geological as well as the topograph- 

 ical conditions may have some influence upon the iiequency of lightning stroke. 

 According to the authority already (juoted,' if 1 represents the frequency of light- 

 ning stroke in a chalk formation, 2 will represent the liability for marl, 7 for clay, 9 

 for sand, and 22 for loam. 



"With regard to trees, the oak is most freque^itly and the beech the least fre- 

 quently struck. The values are something like, if 1 represents the frequency for 

 the beech tree, 15 for pines, other trees generally averaged at 40, and 5i for oaks." 



The different kinds of lightning flashes and the function and value 

 of lightning rods are discussed. 



"The statement that lightning always follows the path of least resistance, as com- 

 monly understood and stated, needs modification. True it is that when the air is 

 strained by being subjected to the electrifications of cloud and earth, the weakest 

 spot gives away first, and this is apt to be in line with some small elevated knob or 

 surface; but it is equally true, and is, perhaps, the more general case, that when a 

 really vigorous disruptive discharge docs occur, it is somewhat, as Dr. Lodge apHy 

 puts it, like an 'avalanche.' As a matter of fact, we find from the study of actual 

 cases where buildings have been struck that lightning often disregards entirely 

 metallic surfaces and points. What we should first know is whetlier the condition 

 is to be one of 'steady strain'- or 'impulsive rush'^ discharge. In the case of 

 •steady strain" the metal is apt to influence the path of discharge; in the case of an 

 'impulsive rush" discharge even points seem to lose their efficiency and become of 

 little use. . . . 



"While not going into details regarding the question of tlie sliape of the rod. let 

 ns emphasize the fact, so recently brought out, that if an electric current flows 

 steadily in one direction in a cylindrical Avire its intensity is the same in all por- 



' Royal Prussian Bureau of Statistics. 

 * Terms used bv Prof. Lodge. 



