16 REPORTS ON THE STATE OF SCIENCE. —1917. 
The work of the Carnegie Institution has two aspects which specially 
call for attention. Its work at sea by the employment of a vessel,,the 
‘ Carnegie,’ practically free from iron, appears to be so superior in accuracy 
to that hitherto done by the different Hydrographic Departments as to 
rather encourage these bodies to leave to the Carnegie Institution the 
business of obtaining the data necessary for the construction of charts. 
There is some risk lest the work of the Carnegie Institution be made 
an excuse for reducing official provision for necessary magnetic work, 
especially survey work. The other aspect, though perhaps less 
intrinsically important, seems more pressing. The Carnegie Department 
of Terrestrial Magnetism naturally aim at a uniform standard in their 
survey work, and they desire to use not merely their own but all available 
field observations. To this end they have carried out a large number of 
comparisons between numerous instruments of their own, and also between 
their instruments and those of many foreign observatories. They have 
recently been reducing their results to what they hope may be accepted 
as an international standard, and this standard has been already adopted 
at several observatories, including those of the U.S. Coast and Geodetic 
Survey. But it is important to remember that this standard is a purely 
American choice, and does not at present embody any formal international 
agreement. It seems not improbable that magneticians may presently 
find themselves in a somewhat similar position to that occupied twenty 
years ago by electricians as regards resistance standards. The increased 
refinements of late years may enable electrical measurements to be made 
with an accuracy justifying six significant figures, but I hardly think a 
5-figure accuracy—which the so-called international standards seem to 
postulate—can yet be satisfactorily claimed for absolute magnetic measure- 
ments. Until this accuracy can be secured, not occasionally and acci- 
dentally, but regularly, no convincing answer can be given to the query 
whether the indications from a so-called standard instrument are unchange- 
able from year to year. 
The construction of absolute magnetic instruments, giving a 5-figure 
accuracy, is presumably merely a question of time and expense. In the 
meantime, we shall probably have to content ourselves with something 
less. Buta good deal might be learned, and a very useful purpose would be 
served, if a workable scheme could be arranged for the systematic com- 
parison of the instruments in use at the magnetic observatories in the 
British Isles. It is clear that any international scheme which may come 
into operation after the war would be much facilitated if each country 
made itself responsible for the intercomparison of all instruments within 
its own bounds. <A satisfactory comparison of instruments, however, 
involves considerable time and expense; thus, unless its necessity is gene- 
rally felt, it is unlikely to be accomplished. 
‘At one time, magnetographs of the Kew pattern were in a considerable 
majority, and absolute instruments—magnetometers and dip-circles— 
were largely made in England and verified at Kew Observatory. Also, at 
one time, a considerable number of foreign, Colonial, and Admiralty 
observers came for instruction to Kew. All these things tended to 
uniformity of practice. Circumstances have, however, greatly altered. 
Large magnets, such as those of the Kew pattern magnetograph—in their 
day representing a great reduction from the magnets of the Gaussian era— 
have gone out of fashion, partly for substantial reasons, and partly because 
