Magnetic Instruments on the Galilee 23 



obtaining values of the magnetic declination within 2' or 3'. The instrument is not recom- 

 mended, however, for general land work, having been designed to meet the special needs of 

 work at sea. 



In addition to perfecting the instrument itself, special experiments have been in progress 

 with the view of disclosing the cause of outstanding errors. There is no great difficulty in 

 perfecting a magnetic instrument which shall admit of observations with the desired 

 absolute accuracy over a lunited region, but when the same accuracy must be insured 

 over practically the entire globe, then problems present themselves not readily appreciated. 

 Even for land instruments, as has been repeatedly found, the problem is not such a simple 

 one. Accordingly, in our work, great stress has been laid on the necessity not to overesti- 

 mate the absolute accuracy obtained, but, on the contrary, continually to assume that the 

 desired accuracy is not being reached and, hence, that it is of the utmost importance to 

 get independent checks in every possible manner. Thus not only has every opportunity 

 been embraced in port to get shore intercomparisons between all ship instruments and 

 land outfits (our own as well as those of local observatories), but there also have been 

 devised special testing-appliances at Washington. 



One peculiarity of the various sea dip-circles which have come to our notice has not 

 yet been wholly explained, viz, well-nigh invariably the dip-needle corrections are negative 

 on good land dip-circles, or on approved earth-inductors. This correction may be as much 

 as 5' and more; hence the need of the continual control spoken of in previous paragraphs. 

 In these instruments the house in which the needle swings is of brass, whereas in land dip- 

 circles it is of wood. Throughout our experience, covering magnetic instruments of every 

 type and make, we have not yet found one — be it a magnetometer or a dip circle — that has 

 proved wholly satisfactory if the magnet house is of brass. Accordingly, one of our future 

 experiments will be to replace the metal house of the sea dip-circle with wood to see whether 

 the rather large absolute corrections can thus be avoided. 



The possibility of a better way of mountmg the needle than in the present jewels is also 

 receiving attention. Furthermore, a marine earth-inductor has been designed and installed 

 on the Carnegie to serve as another means of control on the sea dip-circle. (See pp. 196-200.) 



For further information regarding the sea dip-circles and best methods of observing, 

 see the Carnegie work, pages 195-196, and the extracts from instructions (pp. 115-127). 



SEA INSTRUMENT FOR HORIZONTAL INTENSITY. 



As described in the previous section, the modifications introduced in the original sea 

 dip-circle made it possible to obtain total intensity {F) observations in all magnetic lati- 

 tudes, beginning with the second cruise of the Galilee. The incUnation or dip (/) being 

 observed at the same time with the same instrument, the value of the horizontal intensity 

 (//) is obtained by computation with the aid of the formula 



H = F cos I 



In accordance with our adopted principles, it was highly desirable also to obtain H 

 directly, in regions of not too low values of H, i. e., in not too high magnetic latitudes, by 

 some convenient and independent method. Accordingly in the spring of 1905, L. A. Bauer, 

 having in mind this desideratum and the failures experienced in low magnetic latitudes 

 with the original sea dip-circles, undertook the devising of a special and simple deflecting 

 apparatus, which could readily be attached, if necessary, to the ordinary navigating 

 compasses, or form an entirely independent instrument. At that time Bidlingmaier's 

 "double compass" had not been perfected, and even if it had, it would not have answered 

 our requirements. The simplest possible contrivance was desired both from an instru- 

 mental as well as from a computing standpoint. For one reason or another previous appli- 

 ances for measuring the horizontal intensity at sea had not proved entirely satisfactory. 



