686 
being multiplied. A question requiring 
further consideration is the number of sizes 
of particles—separates—to be recognized 
and the limits of these. Several systems 
of groupings are now in use.2? ‘The finer 
the particles the greater is the influence of 
a given mass of them. upon the character 
of the soil. Many more divisions should, 
therefore, be made in the fine material than 
in the coarse material. It is an open ques- 
tion whether, as survey work has been done 
in the United States, sufficient divisions 
have been made below the sand _ classes. 
Undoubtedly field separation of materials 
by hand examination is not likely to be 
more refined than is possible with the divi- 
sions now generally made. However, finer 
distinctions in the mechanical analysis of 
material smaller than 0.005 mm. may ex- 
plain some variations in types of soil not 
otherwise recognized and these form the 
basis for more detailed study of individual 
types. 
The structure of a soil as determined by 
the order of stratification and the thickness 
of the layers may also be the basis of type 
separation. This is independent of general 
structural differences due to the general 
mode of formation and the characters of 
the rock. 
The soil type is the unit for soil study 
and should be as nearly alike in all parts 
as is possible. It is the most important 
grouping of material primarily because it 
does represent the chief physical differ- 
ences in soils. The next most important 
grouping is the soil series and these two 
will be most generally identified with par- 
ticular crop and agricultural interests in 
practise. This does not minimize the value 
of the larger separations, which, as has been 
suggested, are essential to reasonable accu- 
racy in these last two groups. 
* Briggs, L. J., et al., ‘‘The Mechanical An- 
alysis of Soils,’’ Bul. 24, Bureau of Soils, U. 8S. 
Dept. of Agr., 1904. 
SCIENCE 
[N.S. Vou. XXXV. No. 905 
Of course the final test of a survey must 
be in the field man who applies these prin- 
ciples to a particular set of conditions. 
Owing to the intimate overlapping of seyv- 
eral fields of natural science in this work it 
is evident that he should be a man of broad 
training, including especially geology and 
the principles of soil fertility, and he should 
have keen power of observation and cor- 
relation. Scarcely any experience or train- 
ing which the field man may possess but 
finds use in the ideal soil-survey man. 
EuMer O. FIPPIN 
CoRNELL UNIVERSITY 
THE BRITISH ASSOCIATION 
A PRELIMINARY program has been issued for 
this year’s meeting of the British Association, 
which, as already announced, is to take place 
at Dundee from September 4 to 11. The meet- 
ing will be the eighty-second of the series, the 
twelfth in Scotland, and the second in Dun- 
dee, the association having previously met in 
that city in 1867. 
The opening meeting will be held in the 
Kinnaird Hall on Wednesday evening, Sep- 
tember 4, when Professor E. A. Schafer, 
F.R.S., will assume the presidency and de- 
liver his inaugural address. In the same hall 
the first evening discourse will be delivered on 
Friday, September 6, by Professor W. H. 
Bragg, F.R.S., on “ Radiations Old and New,” 
and the second on Monday, September 9, by 
Professor A. Keith, on “The Antiquity of 
Man.” The reception room and offices will be 
established in the Albert Institute, and a con- 
siderable proportion of the sections will have 
their meeting-rooms in the University College. 
Arrangements have been made with the 
railway companies for the issue to members of 
return tickets at reduced fares, available for 
the period of the meeting and eight days 
after, and excursions during and after the 
meeting, for the purposes of scientific field- 
work, are expected to prove particularly im- 
portant this year. The famous Alpine flora 
of Clova and Glenesk, the fossil fish beds of 
Dura Den, and the geology of the Stone- 
haven region and of the Western Highlands 
