566 
of specimens, and it is here that the recent 
way of studying natural history and anthro- 
pology makes itself felt. Formerly stu- 
dents were satisfied if they could see one or 
two specimens of any given species of plant 
or animal, or an example of a shield or a 
spear of a given tribe. Now they have 
learnt that it is necessary to have long 
series of representatives of a species or of 
the various implements and weapons of 
primitive races. The old order belonged 
to the period when the fixity of species was 
an article of faith. To-day the naturalist 
is more interested in varieties and interme- 
diate forms. Formerly the naturalist de- 
lighted in clearly cut classifications ; now 
he revels in tracing the infinite gradations 
of nature and in endeavoring to learn their 
significance. The same applies to the an- 
thropologist ; those, for example, who study 
the decorative art of savage peoples or of 
pre-historic times have to visit as many 
museums as possible in order to get a suf- 
ficiently long series to enable them to trace 
the origin and distribution of certain de- 
signs. No museum is likely to be too large 
for such students. This is what may be 
termed the dictionary type of museum, and 
the specimens contained within it require, 
like the words in a dictionary, to be so ar- 
ranged that they can be referred to with 
the minimum amount of trouble. 
Those who have seen the great stores of 
duplicates in such museums as the Na- 
tional Museum, of Washington, or the 
Agassiz Museum, at Harvard, will recog- 
nize that our American colleagues have real- 
ized this need ; but there are many practical 
difficulties. In large towns space is too 
valuable to admit of the accumulation of 
long series, unless they are to be so stored 
as to be studied under great difficulties ; 
and certain large objects, such as many an- 
thropological specimens, would occupy so 
much space as to preclude their being col- 
lected. Further, the cost of the dust-proof 
SCIENCE. 
[N. 8. Vou. VI. No. 146. 
cases which are absolutely necessary in 
cities is very great. 
Appreciating the need for the collecting 
of long series of anthropological specimens ; 
for the desirability of having representa- 
tives, or representations, of all the objects 
made by vanishing tribes of natives; and 
for the advisability of keeping together as- 
sociated objects from extensive archzeolog- 
ical excavations, Professor Flinders Petrie 
has devised a scheme which will give the 
greatest possible storage space for a given 
expenditure. This scheme, which he brought 
before the meeting of the British Associ- 
ation at Liverpool, is briefly as follows : 
The conditions for a repository with such 
a scope are so wholly different from those of 
existing museums that the proportions of 
expenditure are entirely changed. The es- 
sential and primary condition is that space 
shall be of minimum value; and, as it is 
desirable to keep down wages and the cost 
of moving objects, it is needful that, what- 
ever the amount of expansion, no rear- 
rangement should be necessary. 
The type of structure must, therefore, be 
a long gallery, with lateral expansions to 
be built as any section increases. The gal- 
leries must be sufficiently wide apart to al- 
low of any likely increase, however irregu- 
larly distributed. 
An economical type of gallery would be 
one about 54 feet wide, divided into a nave 
and two aisles, the latter being subdivided 
into bays, 16 feet long. From these bays 
lateral expansions would be added when- 
ever necessary. The walls should be low 
—say 10 feet—and about one-fourth of the 
roof should be of glass, which would ensure 
the galleries being well lighted. 
The essence of the scheme is that the 
site shall be ordinary agricultural or wooded 
land, so that, however irregular the expan- 
sion, the unoccupied land will continue to 
be productive. Thus every possible need of 
the future can be accommodated without 
