Septembee 27, 1901.] 



SCIENCE. 



497 



most quickly reaches a very sensitive state of 

 instability and perhaps the sun-heated air on 

 the rising slope of a mountain may determine 

 the start of the atmospheric overturn. As the 

 atmospheric collapse spreads over the continent 

 its character may be greatly influenced by the 

 degree of instability existing in the regions 

 over which it passes and by the existence of 

 independent storm movements. Furthermore 

 it is known to be the case that a typical cy- 

 clone in the United States causes a great mass 

 of warm air to be gathered near the earth's 

 surface on its southeast front and when this 

 mass of warm air and the overlying cold air 

 make a summersault a tornado (cyclone, popu- 

 larly called) or severe local disturbance is the 

 result. 



Our dominoe storm, to carry our analogy 

 further, might be inaugurated with indefinitely 

 small effort at a time when the system is ready 

 for a more or less complete collapse, and the 

 trend of the collapse could be controlled not 

 only by choice of time and place of starting the 

 collapse, but also by starting independent 

 collapses at other times and places, and the 

 control of weather must likewise consist of 

 proper starting of storm movements and of 

 their proper modification by independently 

 inaugurated movements. 



Reports are coming to us from southern 

 Europe of the control of hail storms by means 

 of a special form of cannon which throws a 

 large vortex ring at high velocity into the upper 

 atmosphere. In many details these reports are 

 absurd, while in other details they are by no 

 means absurd, although it must be admitted, if 

 we credit the reports, to be a very remarkable 

 fact that this first crude trial to control the 

 weather — for it is the first that conforms at all 

 to the physical requirements of the case — should 

 be in so large a measure successful. 



The problem is to upset the increasing in- 

 stability of the atmosphere on a hot summer's 

 afternoon before the beginning of that particu- 

 lar type of collapse, whatever it may be, that 

 constitutes a hail storm, to set the sky off half- 

 cocked as it were, and it is hard to think of a 

 better means for starting a collapse of an un- 

 stable atmosphere than the smoke-ring cannon 

 of Burgomaster Stiger. A simple concussion or 



loud sound is not at all effective. The thing 

 that is necessary is not a momentary to and fro 

 motion of air such as accompanies a sound wave 

 and which is very slight even in a sound wave 

 of exceedingly great intensity, but an actual 

 ti-ansfer of air from one place to another, such 

 as is produced near the muzzle of a gun in what 

 is called the blast, or such as is produced by a 

 vortex ring. 



It seems to be within the range of possibility 

 that Stiger's cannon may be a means for con- 

 trolling all kinds of storm movements. 



W. S. Franklin. 



REVIEW OF TWO RECENT PAPERS ON BAHA- 

 MAN CORALS. 



To THE Editor of Science : It was my 

 pleasure to visit the American Museum of Nat- 

 ural History in New York during the first 

 week of September, and through the courtesy 

 of Professor Whitfield to examine the recent 

 species of West Indian corals in that institution. 



I saw two specimens that have recently been 

 described by Professor Whitfield, and after 

 having received copies of his papers, desire to 

 make some retnarks on them. 



The first paper is entitled, ' Notice of a Re- 

 markable Case of Combination between Two 

 Different Genera of Living Corals,' Bull. Amer. 

 Mus. Nat. Hist, Vol. XIV., Art. XVII., pp. 

 221, 222, pis. XXXI., XXXII. (date, July 29, 

 1901). Professor Whitfield considers his speci- 

 men a combination of Meandriyia lahyriathica 

 and a Clenophyllia which he says is perhaps 

 nearest to Ctenophyllia quadrata Dana. 



Are tivo genera represented ? Most emphatic- 

 ally no! The Meandrina of Milne-Edwards 

 and Haime (not Lamarck, 1801) is characterized 

 by possessing distinctly toothed septa and a 

 spongy columella, in which may be a lamellar 

 element connecting one calicial center with the 

 next ; the series are variable in length, often 

 very long, and usually sinuous. The wall be- 

 tween adjoining series is simple (not double as 

 in Diploria). The septa and wall are imperfo- 

 rate. Pali may or may not be present ; they 

 are not of specific value in this genus. An ex- 

 amination of plate XXXII. will show that 

 there are no generic differences in the specimen 

 figured. 



