238 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



It will be seen from the table that the mean error is a lit- 

 tle over eight per cent, for the first trials, during which time 

 the instrument (which was the first of its kind) was subject 

 to various sources of error, some of which have now been 

 removed by better mechanical arrangements, and a few of 

 which remain to be overcome. It will also be observed from 

 the second part of the table that in April, 1876, after certain 

 of these mechanical difiiculties had been overcome, the per- 

 centage of error v/as much reduced, so that a sounding by 

 the instrument and one by the line agreed on the average 

 to within ahowt four per cent. Dr. Siemens is now engaged 

 in incorporating some improvements into a new instrument 

 of this kind, and in particular in substituting for the corru- 

 gated diaphragm, which supports the column of mercury, 

 one made of steel which is sawed through in a s|)iral form 

 and covered by a sheet of India rubber. 



In the trial of this first bathometer, Dr. Siemens took it to 

 the top of the great clock-tower of the Houses of Parliament 

 (315 feet), and it was found to read very closely what theory 

 demanded. It may be reasonably hoped that the mechanical 

 difficulties, which are very great, can be so far overcome as 

 to cause the instrument to be read in the open sea within 

 one or two per cent, of the true depth on the average. It 

 must be remembered that the piano-wire apparatus gives the 

 depth of water immediately below the ship, while the ba- 

 thometer gives an indication of an attraction (or of a defici- 

 ency) w^hich is the result of the depth of water for several 

 miles in every direction. It gives the mean depth over a 

 certain area. 



A difficulty will always arise in the use of this device as a 

 navigating instrument in general, as on approaching close 

 to the land it will give an indication due not only to the 

 depth of water beneath the ship, but partially due to the 

 height of the neighboring land above the level of the sea. 

 Thus w^e may expect to find always a reading of the instru- 

 ment near the shore Avhich will be difierent from what it 

 would give in the open sea over the same dejDth of water, 

 and hence, for general cruising, such an instrument will be 

 an uncertain guide of a close approach to land. When, how- 

 ever, the water shoals gradually, so that the 100-fathom line 

 is twenty-five to forty miles from shore in which case the 



