14 



HYDROGRAPHY. 



measurement of the time not being sufficiently minute, and this to- 

 gether with the personal error in observation would, most probably, 

 be found the cause, rather than that assigned for it, viz., the evolve- 

 ment of latent heat by its wave-like motion. On this account 

 I became desirous of repeating the experiments under the most 

 favorable circumstances. An opportunity was afforded by the erec- 

 tion of the Fire Alarm Apparatus recently in Boston, which gave me 

 the opportunity of recording the time on a cylinder, making the 

 second of time equal to about five inches in length, and therefore 

 giving the time with a degree of accuracy that had never been here- 

 tofore obtained. The close agreement of the many distinguished 

 experimenters on sound, had led me to believe that I should but 

 verify the conclusions arrived at, yet I hoped so to vary the experi- 

 ments as to enable full reliance to be placed in the results : these I 

 have embodied in Appendix I, where I have placed them by the 

 approbation of the Joint Committee of the Library of Congress, — in 

 whose name I applied to the Mayor of Boston, the Hon. Benjamin 

 Seaver, for permission to use the apparatus, which was readily and 

 obligingly placed at my disposal. 



The State House at Boston was occupied as the central position, 

 and at other distant stations guns were placed ; while the steeples in 

 which the bells were situated were connected by triangles with the 

 State House, all radiating from it in various directions. The distances 

 were kindly furnished me by Mr. Boutelle, an able assistant on the 

 Coast Survey, from the surveys in progress. After we had arranged 

 this part, it became important that the record of the firing should 

 be made, if possible, by the concussion, or by the same impulse which 

 was given to the air by the discharge of the cannon, so that no per- 

 sonal error might affect the observation, except in the record of hearing 

 the sound. In the case of the bells in the steeples, the magnetic cir- 

 cuit was broken by the hammer the moment it struck the bell ; the 

 personal error then was only to be looked for in the last record to be 

 made, the differences between the two marks on the time-cylinder of 

 course giving the time in seconds and parts of seconds, in lineal mea- 

 surement, the sound had occupied in passing over the distance. The 

 experiments were numerous, — in the greater distances by the report 

 of cannons, and the lesser ones by the bells of the several churches 

 connected with the Fire Alarm Apparatus, both marked by striking 

 the magnetic key. The lines on which the sounds passed radiated in 



