FEBRUARY 23, 1912] 
time, are shown upon graduated circular dials 
on the face of the machine. The machine was 
invented by Dr. R. A. Harris, who seventeen 
years ago published a brief but comprehen- 
sive description and submitted a general plan 
to the Coast Survey Office. (See report of 
superintendent for 1894.) The details of the 
machine-design were worked out by Mr. E. G. 
Fischer, under whose direction the machine 
was set up in the instrument division of the 
survey. 
Mechanical aids of this kind are used in 
connection with the tides, because good pre- 
dictions require the combination of a consid- 
erable number of sine or cosine terms whose 
arguments vary uniformly with the time. 
The new machine contains 32 short-period 
components (7. e., daily, semi-daily, quarter- 
daily, ete.) and 7 having long periods (2. e., 
fortnightly, monthly, semi-annual and annual. 
In combining these numerous terms two 
summations are carried on continuously by 
means of two chains each fixed at one end and 
free to move at the other. Each chain is laid 
alternately over and under a series of pulleys 
whose upward and downward movements 
cause the free end of the chain to travel back 
and forth across a fixed initial point. The 
motion of the free end of one of the chains is 
proportional to the rise and fall of the tide to 
be represented or predicted; that of the free 
end of the other chain (or rather of a marked 
link a certain distance from this end) is such 
that when this link passes across a fixed 
point the height represented by the first chain 
is at its maximum or minimum value. These 
statements describe very briefly the general 
plan upon which the times and heights of the 
tides are mechanically determined. As al- 
ready stated the times and heights are shown 
upon the face of the machine while a curve is 
drawn which makes a permanent record of all 
stages of the tide. 
The machine is driven by hand and the 
gears are such that the periods of motions 
which depend upon them shall represent the 
known periods of the various sine and cosine 
terms into which tidal records or observations 
can be resolved. 
SCIENCE 
307 
In the new machine the error resulting 
from the representation of the incommensur- 
able astronomical ratios by the gears amounts 
to less than one degree for a period represent- 
ing a year in prediction. For the larger 
terms the error is much less, so that after 
predicting a year’s tides and reading hourly 
heights for December 30 and 31, the predicted 
values agreed so well with the values computed 
directly from astronomical data, as to make 
the errors negligible. 
It may be stated that, although tide-predict- 
ing is the most useful purpose to which the 
machine is put, its broadest application is in 
the solution of equations of the form 
y=H,+A cos (at +a) + Beos (bt+8)+---, 
where A, B, O,... denote the amplitudes and 
a, b, c,... their speeds per unit of time f, of 
which it draws the graph and indicates the 
positions and magnitudes of the roots. 
SaMuEL Tierney, JR. 
SPECIAL ARTICLES 
XERALEXIS 
WHOEVER brings forward a new word must 
show what the students in journalism would 
call a “crying need” for it, or take the conse- 
quences. The undersigned is not altogether 
certain as to how crying the need may appear 
to others, but it seems to him that a single 
‘euphonious, appropriate word should be sought 
to replace the clumsy and rather ill-sounding 
compound, “ drought-resistance.” The writer 
proposes the coining of the word szeraléxis, 
from the Greek éyporns, drouth, and aAc€yous, 
a keeping off or resistance. 
We have, of course, the words, “xerophytic” 
and “xerophytism,” which do not, however, 
involve the idea of resistance to drouth in 
their composition, and do not convey that 
meaning in usage, although it stands to rea- 
son that a “drouth-resistant ” plant will have 
“xerophytic”” characteristics of some kind. 
There is ample justification for the new 
word on etymological grounds. The Greeks 
had many compounds in which the above roots 
were employed. The root of the adjective 
aeros, “dry” and xérdtés, “ drouth,” is found 
