1885.] 
Things versus Words. 
517 
will understand us ; but the shape of any objedt, if compli- 
cated and irregular, its colour, its smell, its taste, we cannot 
so describe, save by reference to some other objedt with 
which our hearers or readers are supposed to be acquainted. 
Without some such reference our account of the objedt is 
extremely vague, so vague, in fadt, as rarely to lead to its 
identification. I will take a very simple instance. You tell 
me that some given objedt, or some part of such objedt, is 
red, and you may speak truly. But how am I to know 
which of the hundred shades, commonly known as red, is 
the precise one which you mean ? Is it a brick-red, or a 
rose-red, or a blood-red, a cherry-red, or a poppy-red ? Is it 
a led which refledts the same colour from whatever side 
it is viewed with regard to the light, or does it change 
like the elytra of certain beetles, the abdomen of the fire- 
wasps, or the throats and breasts of some humming-birds ? 
Without an answer to these questions the inquirer can form 
no definite idea of the objedt in question, and without he 
has a knowledge of the standards of comparison he cannot 
comprehend the answer. 
We may lead in chemical text-books that on bringing cer- 
tain substances in contadt, or on heating them to a high 
temperature, there takes place a change of colour, a preci- 
pitate of some given shade, or a flame which, as the case 
may be, is blue, green, purple, &c. But the student who 
meiely reads these statements, or hears them uttered by a 
ledtuiei without adtually seeing the results, forms but a very 
vague imperfedt notion of the fadts. To this point attention 
is now. being diredted. A chemical manual, recently pub- 
lished in America, is illustrated with portions of the adtual 
piecipitat.es produced, cemented upon slips of paper, and 
the transitory changes of colour produced in certain Solu- 
tions by contadt with various reagents are imitated as well 
as possible. 
Again, the student is now often told that in order to get a 
clear notion of any novel readtion, he must not merely* pro- 
duce it with materials of known purity, but he must make 
tiials with substances which give a closely approximating 
result, so that he may note the shades of difference. 
All such instructions convey the admission that language, 
however invaluable, is a very poor, imperfedt, medium for 
retarding and conveying accurate observations. 
I turn from colour to form. If I say that some unknown 
objedt is of a triangular or of an hexagonal shape, everyone 
understands me. But suppose I have before me some 
newly-discovered animal or plant, how can I describe it so 
