196 REPORTS ON INVESTIGATIONS AND PROJECTS. 



navigators. It was necessary to take an average of the data supplied from 

 many different sources in the hope that thereby the effect of the outstanding 

 errors might be reduced to a minimum. Thus, for example, a few years ago 

 an attempt was made by a hydrographic office to improve the existing mag- 

 netic charts with the aid of data obtained on vessels the world over in the 

 course of their regular voyages. Circulars were sent to many thousand ves- 

 sels and many hundred supposedly accurate determinations of the compass 

 direction were received in response and plotted on a world chart. Whenever 

 there were several values for the same region they were found to be so dis- 

 cordant among themselves that the error of even the average results appeared 

 sufficiently large to render it unsafe to make any material alteration in the 

 existing "lines of equal magnetic variation." In brief, to determine possible 

 errors in the present mariners' charts with the appliances and vessels hereto- 

 fore at the disposal of the hydrographic offices would not only be an exceed- 

 ingly expensive undertaking, but a most laborious task as well, with the 

 unsatisfactory result, after all had been done, that there might still be errors 

 due to variable effects of the iron of the modern vessel which were not suffi- 

 ciently eliminated. 



It is possible for the Carnegie, on a single trip of but twelve days, as, for 

 example, the one in October of 1909 from St. John's, Newfoundland, to 

 Falmouth, England, to uncover errors of practical importance to the mariner 

 because of their consistency and prevalence over long distances. These 

 errors, in general, exceed the corrections on account of the secular change in 

 the compass direction for 10 to 20 years. 



The errors found by the Carnegie on her trip to Falmouth, as above re- 

 lated, were confirmed on other parts of her first cruise covering the Atlantic 

 as far south as the Madeiras and the Bermudas. Again, for long stretches, 

 systematic and, hence, cumulative errors were disclosed, the charts showing 

 the compass direction, which are now in actual use, being found in error at 

 times as much as 2 to 2.5 . The chief cause for these errors was likewise 

 discovered, viz, that incorrect allowances, for sufficiently long periods, were 

 made in the construction of the charts on account of the secular changes. 

 The "Complete magnetic results of the first cruise of the Carnegie" were 

 promptly supplied in manuscript to the leading hydrographic establishments 

 and published in the June 1910, issue of the journal "Terrestrial Magnetism 

 and Atmospheric Electricity." The September issue of the same journal 

 gave an account of the special investigations made in the vicinity of Gardi- 

 ner's Bay in 1909 and 1910, as also tabular statements showing the distribu- 

 tion and run of the chart errors revealed by the Carnegie's first cruise. The 

 errors disclosed may be summarized as follows : 



Except for the portion of the cruise from 48.5 ° N., 47 ° W., to Falmouth 

 Bay, and thence to Madeira, all charts show too low west magnetic declina- 

 tion (variation of compass) over the portion of the Atlantic Ocean covered 

 by the Carnegie. While the error is in general less than a degree, it is un- 

 fortunately in the same direction for about 5,000 miles, and hence the result- 



