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



Since, therefore, the effect of the intensity upon the velocity 

 is quite imperceptible in the case of a cannon-shot, a peculiar 

 interest attaches to a recalculation of the most exact experiments 

 that have been made by this means, which, as we have already 

 said, are those of Moll and Van Beek, in order to deduce hence 

 the value of 7, and from that the mechanical equivalent of heat. 



Calculation of Moll and Van Beek's Experiments. 



These experiments were made on the 27th and 28th of June, 

 1823. The two stations were two hills, Kooltjesberg and 

 Zevenboomen, near Amersfoort, 5 [German] miles from Amster- 

 dam. Moll was at the former station, and Van Beek at the latter. 



The signals were made by discharging cannon simultaneously 

 at both stations. All the observations which were not simulta- 

 neous were excluded, so as to eliminate the influence of the wind. 



The data of the observations are — 



Day. 



Number 

 of determina- 

 tions. 



Velocity per 

 second. 



Barometer. 



Thermo- 

 meter. 



Tension of 

 aqueous 

 vapour. 



27 



28 



44 



28 



m. 

 34006 

 339-34 



m. 



0-74475 

 0-74815 



11-16 

 11-00 



m. 



000925307 

 000840465 



The barometer and ' thermometer were read continuously at 

 both stations, and the above values are the means of all the 

 readings. The observations of the barometer are corrected for 

 capillarity and temperature. 



The statement of the temperature for the 28th of June differs 

 by 0*21 from that employed in the previous reduction of the 

 results, in consequence of an error of calculation having been 

 made on that occasion. 



The velocity of sound in dry air, calculated from these results 

 for 0° and a pressure of 0*76 metre, was — 



27th of June .... 332-38 



28th 



331-72 

 33205 



As some corrections were neglected in calculating this result, 

 and the constants employed in the reductions have since been 

 determined with much greater accuracy by Regnault, we may 

 here repeat the whole calculation. Moll and Van Beek found 

 for the length of the basis 17669*28 metres, a value which was 

 arrived at by triangulation founded upon KrayenhofPs measure- 

 ments. I am, however, informed by my highly esteemed teacher 

 Professor Van Rees of Utrecht, that by a recalculation, in which 

 he applied the various equations of condition which had not 



