FEBRUARY 25, 1904] 
NATURE 
403 
that traces of magnesium (one part in 14,000) in mercury 
cause the latter to decompose water and to oxidise rapidly 
in the air at ordinary temperatures. Thorium with less 
‘than a trace of actinium produces an auto-photograph. 
This point cannot be too strongly stressed in the rare 
earth field. One who has wrought with thorium dioxide 
well knows the influence a small amount of cerium has upon 
its solubility. The conflicting statements in the literature 
as to the colours of the oxides of the complexes, neodidy- 
mium and preeseodidymium, cause one to wonder if different 
researchers have had the same hecceity. 
An appeal to the spectroscope is, of course, in the minds 
-of all my hearers. 
Griinwald, in a series of papers on his theory of spectrum 
analysis,’ endeavours ‘‘ to discover relations between the 
spectra and thus to arrive at simpler, if not fundamental 
“elements.’’’ He came to the conclusion that “all the 
‘so-called elements are compounds of the primary elements a 
and b”’ (coronium and helium). Ames (Am. Chem. J., 
xi., 138, 1889), having directed attention to the use of 
vuncorrected data by Griinwald, remarks: ** The concave 
grating gives the only accurate method of determining the 
ultra-violet wave-lengths of the elements; and as a con- 
sequence of not using it, most of the tables of wave-lengths 
so far published are not of much value.” 
Lockyer maintained that the lines of certain brilliant 
substances vary not only in length and in number, but also 
in brilliancy and in breadth, depending upon the quantity 
of the substance as well as temperature (Roy. Soc. Proc., 
Ixi., 148, 183; Chem. News, Ixxix., 145). Being unable to 
decompose the elements in the laboratory, he studied the 
spectra of the stars. The spectra of the colder stars show 
many more metals, but no metalloids, whereas the coldest 
stars show the Crookes spectrum of metalloids- which 
are compounds; none of the metalloids are found in 
ithe spectrum of the sun. More than 100,000 visual 
observations and 2000 photographs were made in the 
researches. 
Without doubt the spectroscopic criteria are the most 
valuable we have in judging finally the elements, and 
mayhap will remain so, but in my humble opinion, such 
have not alone sufficient authority, as yet, to usher the 
aspirant to a place among the elect. The contention frames 
itself, however, in an expression of the need for uniformity. 
Whether we follow the most advanced metaphysico- 
chemical teachings or no, if there be any one concept upon 
which modern practical chemical thought depends, it is the 
law of definiteness of compositton. There may be, and 
doubtless are, definite, perhaps invariable, properties of our 
elements other than their combining proportions, the atomic 
weights, if you please, yet, so far as we know, they approxi- 
mate more closely than any fixed, if not permanent, ratios. 
Many of these values, by which we lay such store, are de- 
pendent upon data in which, I venture the assertion, too 
great confidence has been bestowed, or opinions to which 
sufficient attention has not been given. 
As hinted at in the earlier portion of this unduly prolonged 
address, many have theorised as to the ultimate composition 
of matter. The logic of Larmor’s (Phil. Mag., December, 
1897, 506) theory, involving the idea of an ionic substratum 
of matter, the support of J. J. Thomson’s (Phil. Mag., 
‘October, 1897, 312) experiments, the confirmation of Zee- 
> : 
mann’s phenomenon, the emanations of Rutherford, 
ay 
Martin’s (Chem. News, |xxxv., 205, 1902) explanations, 
cannot fail to cause credence in the correctness of 
Crookes’s idea of a fourth state of matter (Phil. Trans., 
ii., 1881, 433). In the inaugural address as president of the 
British Association (1898), he acknowledges in the 
mechanical construction of the Réntgen ray tubes a sugges- 
tion by Silvanus Thompson to use for the antikathode a 
metal of high atomic weight. Osmium and iridium were 
used, thorium tried, and in 1896 Crookes obtained better 
results with metallic uranium than platinum. 
These and the facts that most of the elements with high 
atomic weights, in fact all above 200 (thallium not reported 
1 “Uber das Wasserspectrum- das Hydrogen. und Oxygenspectrum,” 
Phil. Mag. Xxiv., 304, 1887. *‘ Math. Spectralanalyse des Magnesiums und 
der Kohle.” Monatshefte fiir Chemtie, viii., 650. “‘ Math. Spectralanalyse 
des Kadmiums,"” Monatshefte fiir Chemtic, ix., 956. 
NO. 1791, VOL. 69] 
| that extent geographic. 
on), exhibit radio-active properties, are doubtless closely 
associated and have to do with the eventual composition of 
matter. I have unverified observations which go to show 
the existence of at least one element with a very high atomic 
weight. If it be confirmed, then we have them now or they 
are making, and probably breaking up, as shown by that 
marvellous class of elements in the discovery of which the 
Curies have been pioneers. 
If our ideas that all known elements come from some 
primordial material be true, then it stands to reason that 
we are coming in time, perhaps, to that fixed thing, a frozen 
ether, the fifth state of matter. I may make use of 
dangerous analogy and liken our known elements, arranged 
in a perfected natural system, to the visible material spec- 
trum, while electrons, &c., constitute the ultra-violet and 
cosmyle composes the infra-red, either one of the latter by 
proper conditions being convertible into perceptible elemental 
matter. 
Tue Scope oF GEOGRAPHY.” 
The essential in geography is a relation between the 
elements of terrestrial environment and the items of organic 
response ; this being only a modernised extension of Ritter’s 
view. Everything that involves such a_ relationship is to 
Anything in which such a relation- 
ship is wanting is to that extent not geographic. The loca- 
tion of a manufacturing village at a point where a stream 
affords water-power is an example of the kind of relation 
that is meant, and if this example is accepted, then the 
reasonable principle of continuity will guide us to include 
under geography every other example in which the way 
that organic forms have of doing things is conditioned by 
their inorganic environment. The organic part of geo- 
graphy must not be limited to man, because the time is 
now past when man is studied altogether apart from the 
other forms of life on the earth. The colonies of ants on 
our western deserts, with their burrows, their hills, their 
roads and their threshing floors, exhibit responses to 
elements of environment found in soil and climate as clearly 
as a manufacturing village exhibits a response to water- 
power. The different coloration of the dorsal and ventral 
parts of fish is a response to the external illumination of our 
non-luminous earth. The word arrive is a_ persistent 
memorial of the importance long ago attached to a success- 
ful crossing of the shore line that separates sea and land. 
It is not significant whether the relation and the elements 
that enter into it are of easy or difficult understanding, nor 
whether they are what we call important or unimportant, 
familiar or unfamiliar. The essential quality of geography 
is that it involves relations of things organic and inorganic ; 
and the entire content of geography would include all such 
relations. 
Thus defined, geography has two chief divisions. Every- 
thing about the earth or any inorganic part of it, considered 
as an element of the environment by which the organic in- 
habitants are conditioned, belongs under physical geography 
or physiography.* Every item in which the organic in- 
habitants of the earth—plant, animal or man—show a re- 
sponse to the elements of environment, belongs under 
organic geography. Geography proper involves a consider- 
ation of relations in which the things that belong under its 
two divisions are involved. 
Geography is, therefore, not simply a description of places ; 
it is not simply an account of the earth and of its inhabitants, 
each described independently of the other ; it involves a rela- 
tion of some element of physical geography to some item of 
organic geography, and nothing from which this relation 
is absent possesses the essential quality of geographical 
discipline. The location of a cape or of a city is an elemen- 
tary fact which may be built up with other facts into a 
relation of full geographic meaning ; but taken alone, it has 
about the same rank in geography that spelling has in 
language. A map has about the same place in geography 
1 See the exquisite paper by Madame Curie on ‘‘ Radio-active Substances,” 
also “‘ Radio-active Lead,” Hofmann and Strauss, Berichte i 
Pellini (oc. cit.) on “‘ Radio-active Tellurium” ; Strutt, P% 
113, Elster and Geitel, Giesel, Marckwald, *c. 
2 Abridged from an address delivered before the Section of Geology and 
Geography of the American Association by Prof. W. M. Davis 
3 It should be noted that the British definition of physiography gives it a 
much wider meaning than is here indicated. 
*y 30333 
@. Mag., Vi.y 
