August .8, 1891.J 



SCIENCE. 



123 



We are now prepared to investigate the value of the telegraphed 

 result from Texas. Any one who will examine the weather maps, 

 now sown broadcast over almost the whole country, will find that 

 OQ the 11th instant there was a natural rain which extended over 

 the whole of Texas and adjacent regions. One thing seems very 

 evident, that absolutely no rain can be obtained out of a dry atmos- 

 phere. If the explosions can produce rain in limited quantities, 

 yet their influence must always be exceedingly slight, and the 

 expense of the explosions must always be all out of proportion to 

 the amount of good done. Professor Harrington has well said that 

 these experiments begin at the wrong end. The time may be ripe 

 for experimenting in the atmosphere upon the cause of rain, about 

 which we now know practically nothing. It must be conceded 

 •that until we do first experiment upon the cause of rain, all time 

 and money used in making gross explosions will be wasted. 



H. A. Hazen. 



Washington, D.C., Aug. 17. 



P. S. — Since writing the above, a telegram from Midland, 

 dated Aug. 19, states that several more preliminary explosions 

 were made on Aug. 18, and that immediately thereafter rain be- 

 gan falling and continued over four hours. An examination of 

 ihe weather maps for Aug. 18 has shown that the rain began, to 

 the north of Texas, at least eleven hours before the explosions, 

 and covered an area of over 800,000 square miles. The final 

 tests were to be made on the 20th. H. A. H. 



Experiments on Snake Locomotion. 



It is a well-known fact that a snake moves along over the ground 

 by means of adjustable plates or scutes situated on the ventral 

 surface of the body. How the movements of these scutes succeed 

 each other, and what relations the different convolutions of the 

 body bear to one another, are not so satisfactorily known. Who- 

 ever has examined the mechanism of the scutes will, I think, come 

 to the conclusion that they must be moved by the costal muscles, 

 and that this movement must consist in a posterior depression by 

 which tlie scute offers an opposing surface to the ground. In all 

 probability this depression is both downward and backward, thus 

 imparting a slight forward impulse to the body. If this view of 

 the case is correct, we would naturally expect that the act of loco- 

 motion would consist in some sort of fusion or succession of these 

 minute individual impulses. Owing to the rapidity with which 

 these movements are normally executed it is impossible to analyze 

 or define their exact nature, and accordingly experiment seems 

 to offer the only trustworthy guide to a solution of the problem. 

 In experimenting, however, we are encountered with great diffi- 

 ■culties at the very outset. 



If we could succeed in recording the movements of an animal 

 by means of apparatus, the construction of which was ever so 

 delicate, can we rely on this record as a faithful expression of the 

 natural and unimpeded movements of the animal? Wecanhardly 

 feel at liberty to do so. There are at least two causes which may 

 vitiate the results: (1) the animal is excited and annoyed by the 

 experiments and does not act naturally ; (3) the apparatus used in 

 the experiment may directly impede the organs in the discharge 

 of their normal functions. But while these difficulties render it 

 impossible to obtain a record which is trustworthy in all respects, 

 jet approximate results may be obtained which will lead up to a 

 correct solution in the end. 



In considering the locomotion of the snake, it may be well first 

 to state what we know and what we do not know. We know that 

 the snake generally moves on a horizontal or inclined plane, rarely 

 elevating any part of the body to a very considerable distance above 

 that plane. It sometimes moves with its body straightened and 

 in a straight line, but far more frequently the body is placed so as 

 to resemble a sinusoid, and its movements have a lateral and a 

 direct component. The larger convolutions of the body occur in 

 those portions which have the greatest mean diameter. The con- 

 volutions do not form simultaneously, but each travels the whole 

 length of the body, like a wave of water, being at no two consecu- 

 tive moments composed of the same parts. These waves succeed 

 «ach other on opposite sides of the body, thus producing a recipro- 

 cal xjurve. Each wave travels from the head towards the tail, and 



drives its predecessor of opposite phase before it until it disappears 

 at the tail. At times the curves do not shift to alternate sides of 

 the body, but successive curves are formed on the same side. This 

 motion, be it observed, is totally distinct from the reciprocal curv- 

 ings described above. So much for what can be directly observed. 

 But we cannot tell by direct observation the curves which different 

 parts of the body would describe were they to mark the surfaces 

 over which they move. Nor can we observe the movements of 

 the scutes, or their correlations with the movements of the body 

 as a whole. If we are to understand these activities, we must do 

 so by experiment. 



The following was the method of experiment employed. Short 

 pieces of thread were run through bits of sponge saturated with 

 ink, and these were tied around the body of the snake so that the 

 sponges would come on the ventral surface. When these were 

 securely tied the animal was placed on a strip of coarse paper and 

 allowed to move. So long as the sponges were properly supplied 

 with ink every movement made by the parts of the body thus 

 provided was marked on the paper. Now if the different sponges 

 were soaked with. ink of different color, the simultaneous move- 

 ments of different parts would be recorded, and, theoretically, 

 with a sufficient number of sponges placed at proper intervals we 

 would secure a complete record of all the bodily movements 

 during a sustained period of locomotion. Such a record, how- 

 ever, it is impossible to obtain, for reasons which need not be men- 

 tioned. 



The curves obtained by this method were by no means uniform, 

 but varied both with the direction and velocity of the movements, 

 and apparently with the caprice of the animal. The separate curves 

 described by different parts of the body cannot be said to be char- 

 acterized by any marked idiosyncracies. On the contrary, they 

 appear to vary at random, now being marked by acute angles fol- 

 lowed by beautifully rounded sinuosities, which in turn may be 

 succeeded by protracted and irregular curves or at times figure-of- 

 eight tracings. There is this distinction, however, between the 

 curves described by the middle of the body and those of the distal 

 parts. They have not so great an amplitude and are less variable. 

 Contrary to what we would naturally expect, the synchronous 

 curves described by different parts of the body have no discovera- 

 ble agreement either in phase or in form. 



From this description it might be inferred that very little of 

 value could be derived from a study of such curves. But further 

 study shows this inference to be ill-sustained. In interpreting the 

 curves it is well to remember that they do not represent perfectly 

 normal movements, because the scutes over which the sponges 

 were tied were impeded in their action, and because rough paper 

 is even smoother than the average ground over which the snake 

 moves. Owing to this last circumstance the scutes would slip, 

 and the curves would thus be shorter. 



After making due allowance for the conditions which embarrass 

 the experiments, we may perhaps still speak with some degree of 

 confidence as to the general results, and possibly discover the ex- 

 istence of some fundamental laws. Perhaps the most striking 

 fact about all the curves is, that, with very rare exceptions, they 

 are described on opposite sides of an ideal line which may be 

 called the axis'of motion. While they demonstrate that the snake's 

 body is capable of an almost infinite variety of movements, yet 

 lateral movements generally prevail. There is also a tendency to 

 consecutive repetitions, sinuosities following sinuosities, and angu- 

 larities following angularities. The most irregular curves are de- 

 scribed when the animal executes slow and hesitating movements. 

 In this case the curves may be extended on both sides of the axis 

 of motion, or confined to one side, when the curve is a tolerably 

 regular succession of semicircles whose adjacent arcs form cusps. 

 During rapid motion the sinusoid is by far the most common curve 

 described. In fact, it may be regarded as the typical curve de- 

 scribed by the snake's body. 



It is instructive to note that when the curve assumed by the 

 body is a sinusoid, then the curves described by different points of 

 the body are sinusoids. The relation becomes intelligible when 

 we reflect that the curves of the body partake of a wave- like mo- 

 tion, each particle vibrating, as it were, from the crest of one con- 

 volution across the axis of motion to the crest of a succeeding 



