SEPTEMBER 1, 1899. ] 
must of necessity be acquired in addition to 
this engineering education, except when 
such work is in some one of the engineer- 
ing branches. I desire also to call your 
attention to the fact that I do not mean to 
say that it is the only true method to gain 
the desired end, but that I have the feeling 
that, although all roads lead to Rome, yet 
this one is perhaps the most direct and pos- 
sibly the best paved. 
The object of scientific research work is, 
as I understand it, to ascertain the facts of 
nature, to correlate these facts, and finally 
to deduce the laws of nature as illustrated 
by the facts discovered. It is probable that 
a better definition might be given, but for 
the purpose of this address it is desirable 
that the various objects of scientific re- 
search work be given in something like the 
form above, and I feel very confident that 
the definition is one against which no seri- 
ous objection can be raised. It will be 
noticed that I divide scientific research 
work into three parts, and I am sure that 
everybody will agree when I say that 
most of the scientific work done to-day is 
along the first line. The work of correla- 
ting the facts discovered is less common, and 
drawing conclusions from the facts thus cor- 
related by establishing the laws of nature is 
of such a character that but few scientific 
workers get so far. That this is the case is 
but natural, because of the difficulties of the 
problem, and, although it is the most impor- 
tant of all scientific research work, yet, even 
of those who work in this special line, there 
are but few who are able to draw the cor- 
rect conclusions, and this because of the 
lack of proper judgment in weighing the 
importance of supposed evidence and facts, 
or because of lack of previous training in 
suppressing the natural tendency to over- 
estimate the value of one’s own labor. I 
take it for granted that everybody appreci- 
ates the difficulties and failings which, be- 
cause of the previous preparation, or, per- 
SCIENCE. 283 
haps better, because of the lack of previous 
preparation, necessarily attach to the sci- 
entific research work of to-day. Aninquiry 
into the necessary and desirable qualifica- 
tions for a man working in this line will, I 
think, most rapidly lead us to the heart of 
the question raised in this address: Is 
not engineering education a remedy for a 
good many of the weaknesses found in the 
ordinary scientific man ? 
These qualifications will here be given 
separately as effecting the three kinds of 
work into which I have divided all scien- 
tific research work: First, what should be 
the previous training of a man who is to as- 
certain the facts of nature? He must neces- 
sarily have his faculties of observation 
trained to the highest degree, so that he 
sees the facts as they actually are, and, per- 
haps more important yet, that he can see 
the single fact which he is looking for, 
without being disturbed by surroundings 
which, in the eyes of the untrained man, 
would obscure the perhaps small objects for 
which the investigator is looking. It is 
also very necessary that the observer should 
be able accurately to describe the object, or 
fact seen. His mental habit should be 
such that accuracy is a necessity. And, 
finally, I think that the scientific worker 
should, for many kinds of observation, 
have his hands trained in such a manner 
that necessary apparatus can be used in- 
telligently and even designed and made. 
Second, what are the special qualifications 
for collecting and correlating the: isolated 
facts of nature, as discovered by others? It 
seems to me that if anybody is to do this 
successfully he must possess all the qualifi- 
cations of a worker in the first line, ex- 
cept that possibly he does not need the 
manual dexterity which is required by 
the original investigator. In order prop- 
erly to classify the facts according to already 
existing rules and laws, he must also be 
familiar with the methods by which the 
