48 DISCOVERY REPORTS 



CORRECTION OF ECHO SOUNDINGS 



As was stated in our earlier report, echo sounding saves a tremendous amount of time, and at 

 moderate or great depths it is more accurate than sounding by wire. The tables compiled for the 

 Hydrographic Department by Matthews for correcting the speed of sound in sea water have been 

 much amplified since our previous report, as a result of the acquisition of considerably more hydro- 

 logical data, and the new edition of 1939 contains a more accurate delineation of the increased number 

 of areas for which correction tables are now available. As this new information was not available for 

 use on board the ' Discovery II ' before the end of her last commission it has been of interest to note 

 that the original tables in use on board, most of which had been compiled from our own hydrological 

 observations, by the methods shown in the first edition (1927) of Matthews's tables, vary only slightly 

 from the more recent publication. The difference after correction by the new tables is in most instances 

 so small (usually not more than i or 2 fm. in 2500 fm.) that it has not been considered worth while 

 to re-calculate the many thousands of soundings involved, especially as an error of i , 2 or even more 

 fathoms may easily occur, either in reading off the soundings direct from the recorders or from the method 

 used for timing the echo of the old 'listening' receiver. In the latter instance it was not possible at 

 the greater depths to read to a greater degree of accuracy than ± 0-005 sec. — equivalent approximately 

 to an error of ±2 fm. This accuracy, however, was difficult to attain and it is probable that the 

 minimum error was more nearly ±4 fm. Since the accuracy of the soundings also depended greatly 

 on the speed of the machines or recorders, periodic checks were made with an accurate stopwatch, 

 and a rating error noted if the speed was greater or less than the normal. An error, however, which 

 was difficult or almost impossible to detect arose in reading off the depths from the recorders. This 

 reading off can be done in two ways : (a) The depths can be read directly off the wet paper as it passes 

 the scale on the recorder, or (b) the paper can be dried and the soundings read off at any time after- 

 wards with a special scale for the dry paper. Each method has its advantages but it was our practice 

 to use method (a) and enter the soundings in a special log-book. This enabled a complete check to be 

 kept on the exact time of the sounding, since the minute markings on the records could be compared 

 with a chronometer watch, and had the added advantage that an operator was always present when 

 survey work was in progress, to report any sudden irregularity or approaching danger. Method (b) 

 on the other hand did not require such a constant watch to be kept on the recorders, but had the 

 serious disadvantage in the 'Acadia' recorder that whether the paper was dried artificially by the 

 heater incorporated in the lower part of the tank face, or was allowed to dry naturally away from 

 a strong light, the shrinkage of the paper on drying was not always even. This was especially so when 

 the paper was dried artificially, and as a final check on the soundings was always required when they 

 were used for survey work, the heater was disconnected and the paper allowed to dry naturally before 

 storing. Even then there was the disadvantage that the final width of the dried paper might not 

 always be exactly the same, on account of the varying humidity and temperature of the air. 



When soundings are being obtained on moderate or steep slopes it is, theoretically, possible to 

 correct for the errors thus set up. In practice, however, the problem is one of great complexity, and 

 will be discussed at length later in this report. It is sufficient here to say that in general it has not 

 been considered practicable to correct our soundings for slope error. 



