July i8, 1890.] 



SCIENCE. 



33 



taiued in tornado prediction, but was due primarily to an 

 injudicious system of predicting, and secondarily to an im- 

 proper estimate of the nature of the problem. It would be 

 impossible, of course, to say that such a tornado as that at 

 Grinnell in 1882, and the recent one at Louisville, would 

 ■occur in any district. All that we can do is to predict a 

 disturbed region. In verification, it would hardly be fair to 

 adopt principles which could be used in determining the 

 skill of a marksman shooting at a target, for example; but 

 we must take into account the knowledge we have already 

 gained of the relative violence and the manner of occurrence 

 of such storms". We must determine, on a scale, the num- 

 ber of violent storms occurring in any district where such 

 storms were predicted, and not confine attention to the most 

 violent alone. To draw an imaginary line, and say that if 

 a storm occurs within five miles of that line, in a district 

 ■where it was predicted, it shall count fully as a success, but 

 if it occurs five miles on the other side of that line it shall 

 count as a total failure, is to impose restrictions upon the 

 problem which seem entirely unreasonable. 



In a study of tornado predictions made by Mr. Finley for 

 June, 1885, the present writer assumed "that violent storms 

 occurring, in any district predicted for, half way between the 

 centre and edge, shall have weight 1 ; in the rest of the dis- 

 trict, I; to the centre of the next outlying district, ^; to the 

 «dge of that district, ^ ; all outside of these, 0" (see American 

 iTournal of Science, August, 1887, p. 129). The percentage 

 of skill attained as thus measured was 49. Mr. Curtis, tak- 

 ing the same predictions and discussing them mathemati- 

 cally, found 14 per cent. Mr. Curtis has more recently 

 (1887) adopted somewhat the method suggested above, and 

 obtained 40 per cent. These percentages, however, mean 

 very little as to showing a real knowledge of the probable 

 occurrence of tornadoes, for it is necessary to radically 

 change the system of predicting. It would seem wiser to 

 determine as nearly as possible the central point of any 

 probable disturbance, three hundred or four hundred miles 

 to the south-east of a general storm, and then give boun- 

 daries more or less definite to the violence of the storms. 

 This we are able to do from what is known of the behavior 

 of such storms. In verifying, we should consider all the 

 storms that occurred, and give "weights corresponding to 

 their distance from the centre of the disturbed region, and 

 to their intensity. 



Tornado Photographs. 



One of the most recent developments in tornado studies 

 lias been a strong desire to photograph this extraordinary 

 appearance. It is very unfortunate that this desire has be- 

 come so strong that unscrupulous persons have resorted to 

 photographing sketches of tornadoes, and selling them for 

 the real article. It is also unfortunate that all these alleged 

 photographs have been made at distances of from ten to 

 twenty miles. It is a great desideratum that we have many 

 photographs taken at much closer quarters, and this is not so 

 impossible as might at first sight appear. It would be use- 

 less for any one to attempt a photograph on the south side 

 of a tornado within a thousand or fifteen hundred feet; but 

 on the north side we have repeated authentic observations of 

 persons who stood within one hundred and fifty feet, and 

 did not feel any violent wind. It is much to be hoped that 



a photographer will catch, by his instantaneous flash, one of 

 these monsters as it passes just south of his position. It will 

 require more than the usual amount of bravery to do this, 

 however, as is very plain. 



Alleged Photograph. 



While nearly all these photographs show quite plainly 

 their origin, yet there is a single exception in a picture rep- 

 resenting an alleged tornado near Jamestown, Dak., on 

 June 6, 1887, recently published in a prominent magazine. 

 There is no doubt that this is a genuine photograph. There 

 exist most serious difficulties in regarding it a tornado-cloud, 

 however. The picture shows a dense mass of cloud extend- 

 ing from the trees at the earth up to the uniform veil of 

 cloud above, with clear sky on either side. This mass has 

 a thickening on the right-hand side, and this is supposed to 

 be the tornado. The appearance is exactly that of a cloud- 

 burst, as has been often witnessed, and not at all of a tor- 

 nado. The dimensions of the camera and the distance of 

 the cloud give the height between two and three miles. The 

 distance of the cloud was variously estimated from eighteen 

 to twenty miles. There was no destruction, and no one 

 saw it, at the spot where the tornado was supposed to be. 

 The only way it could be located was by following two lines 

 of sight of persons from ten to fifteen miles away until they 

 crossed. Drawings of a sand-whirl, not far from the alleged 

 tornado, showed a funnel-cloud, and nothing at all like this 

 indefinite mass in the picture. The evidence is quite con- 

 clusive that on this day there were in this region several 

 appearances simulating cloud-bursts, tornadoes, and sand- 

 whirls. It is very probable that this photograph was that of 

 a cloud-burst within two or three miles of Jamestown. It 

 is highly improbable that either a cloud-burst or a tornado 

 ever had a height exceeding two or three thousand feet. A 

 photograph of a funnel-cloud showing details, and especially 

 two or three photographs taken as the cloud comes up and 

 passes by, would be of the highest interest, and invaluable 

 at this stage of our studies. H. A. Hazen. 



NOTES AND NEWS. 



The Norwegian Storthing, by 73 votes against 39, has voted a 

 grant of 200,000 kroner for Dr. Nansen's north pole expedition, 



says Nature. 



— The third international shorthand congress will be held at 

 Munich from Aug. 7 to 17, says Nature. The centenary of F. X. 

 Gabelsberger, the originator of modern German shorthand, will 

 be celebrated by those who attend the meetings, and a bronze 

 statue of him will be unveiled. 



— The Entomological Club of the American Association will 

 meet at 9 a.m., on Wednesday, Aug. 20, in the room of Section 

 F, State House, where members of the club will register and obtain 

 the club badge. The president is Professor A. J. Cook, Agricultural 

 College, Mich. ; secretary, F. M. Webster, Lafayette, Ind. Mem- 

 bers of the club intending to contribute papers will send titles to 

 the secretary. The Botanical Club will hold a meeting, as usual, 

 on Thursday, Aug. 21, at the State House. Communications 

 should be sent to the president, Dr. N. L. Britton, Columbia Col- 

 lege, New York, or to the secretary, Dr. Charles R. Barnes, Uni- 

 versity of Wisconsin, Madison, Wis. The Society for the Pro- 

 motion of Agricultural Science will hold its eleventh annual meet- 

 ing in Indianapolis^ beginning on Monday evening, Aug. 18, in 

 the room assigned to Section I in the State House, and continuing 

 on Tuesday. For further information address Professor W. R. 

 Lazenby, secretary, Ohio State' University, Columbus, O. The 



