Dr. Schroder van der Kolk on the Velocity of Sound. 35 



We need, therefore, examine closely only the second method. 



Of all the experiments made by this method, those instituted 

 by Moll and Van Beek in Holland, in 1823, are unquestionably 

 the most accurate. The results obtained by the French Acade- 

 micians in the year 1822, between Monthlery and Villejuif, are 

 not at all to be compared with them. On the first night (that 

 of the 21st of June, 1822) twelve shots were heard at Month- 

 lery and only seven at Villejuif, and on the following night only 

 one shot was heard at Villejuif, wherewith the experiments were 

 ended. 



The signals moreover were not made simultaneously at the two 

 stations, but every five minutes a shot was fired at one or other 

 of the stations. It was thus impossible to apply the proper cor- 

 rection for the influence of the wind ; this can only be done when 

 the cannons are fired simultaneously at both stations. 



The velocity found was 331*2 metres per second, which does 

 not differ much from the result of the later experiments. Never- 

 theless this determination has much less weight than that of 

 Moll and Van Beek. Since this time no similar experiments 

 have been carried out on so extended a scale. 



Parry and Ross in the north-polar regions, and Bravais and 

 Martins on the Faulhorn in Switzerland, obtained tolerably con- 

 cordant results. Those of the former, however, are not very 

 accurate, and there are at least some corrections which cannot be 

 applied with perfect precision to the latter. 



Hence we may confine ourselves to the consideration of Moll 

 and Van Beek's determinations. Although their experiments 

 were performed with the utmost care, the result which they 

 arrived at cannot be accepted, in consequence of the coefficient 

 of reduction not having been accurately known at that time, and 

 of their having, in addition, neglected some corrections in their 

 calculations. Considering, however, the accuracy of the experi- 

 ments, it appeared to me worth while to repeat the calculations 

 with as much accuracy as possible. 



c 

 I was nevertheless convinced beforehand that the value of - , 



c * 

 and consequently that of the mechanical equivalent of heat, could 



not be accurately determined by means of the velocity of sound 

 thus ascertained. In support of this opinion, the following con- 

 sideration may be urged. It is not improbable a priori that 

 these intense air-waves travel quicker than weaker waves ; and 

 the statement of Ross, that in high northern latitudes, in the 

 experiments on the velocity of sound, the report of the gun 

 was heard at the further station before the command to fire, 

 appears to confirm this supposition. If this be the case, deter- 

 minations made by firing a cannon would give a greater velo- 



D2 



