woe 
»4 
724 SCIENCE. 
per ton, is only eight cents per thousand feet (which 
suggests that even our ordinary gas companies make 
profits), but its use is immensely more convenient; 
no stacks are needed, and the furnace reduces to a 
simple non-conducting chamber. The gas has just 
been turned on to the city water-works; and on the 
afternoon of May 22 a well was reported on the prop- 
erty of Mr. Westinghouse, near Pittsburgh. On the 
first day’s excursion numerous furnaces were seen 
running with gas blown in through rough, one-eighth 
inch nozzles; and two or three lines of five-inch pipe 
lay on the surface of the railway embankment. 
Mr. J. W. Cloud, engineer of tests for the Penn- 
sylvania railroad, read a paper on helical springs. It 
was here claimed that round steel is better than 
square, flat, or other shaped; and an investigation, 
mathematical and experimental, was described, on the 
usual and mainly correct hypothesis that the strains 
are entirely torsional. Bars of oil-tempered and un- 
tempered steel, five feet long by three-fourths to one 
and five-sixteenths inches diameter, had been tested, 
and the constants of elasticity, etc., obtained; after 
which the springs had been coiled and again tested, 
and the results compared with theory. The proper 
arrangement of springs, when several are used to- 
gether, was discussed, and certain proportions shown 
to be necessary for springs arranged concentrically. 
Detail drawings of springs for classes V and X were 
shown. Experiment has proved the principles to be 
correct on which these have been designed. Alto- 
gether, the paper is valuable as the commencement of 
an investigation, which, pushed to completion, will 
render the designing of all kinds of helical springs an 
exact science. In the discussion it appeared that 
springs of peculiar shape found their way into the 
scrap pile; that the introduction of peculiar designs 
under freight-cars often resulted in an enormous per- 
centage of breakage; that orders to manufacturers 
are often arbitrary, and contrary to sound principles; 
that logs are loaded on cars by dropping them from a 
height of ten feet; and that springs are tested by 
pounding them together with a steam-hammer, after 
which they are expected to stand ordinary wear. 
The greatest scientific interest, however, attached 
to the paper of Prof. W. A. Rogers of Cambridge, 
on a practical solution of the perfect-screw problem. 
Professor Rogers prefaced the reading by remarking 
that he considered the American society of mechani- 
cal engineers the most appropriate body to receive 
his first public announcement of success, —a cour- 
tesy appreciated by the society. Mechanism of pre- 
cision was defined as perfect ‘‘ when it meets all 
the requirements of the purpose for which it is 
constructed;’’? and the two screws, which raise the 
cross-head of an iron-planer, were discussed in this 
respect. Precision-screws are tested, not only by 
direct measurement of the pitch, but by examining 
optically a surface ruled with many thousand lines 
to the inch by means of the screw. The first catches 
all accumulated errors, while the ‘ diffraction grating’ 
tests the regularity of the spacing for short distances. 
Scales graduated in Europe, and advertised as with- * 
out sensible error, are shown, under the comparator, 
[Vor. IIL, No. 71. 
to merit no such claim: indeed, if we except Pro- 
fessor Rowland’s, no screw has hitherto been made, 
capable of producing graduations sufficiently exact. 
Three half-metre screws were exhibited which could 
be mounted for microscopical examination: on one 
of them, over twelve hundred hours had been spent 
to make it, by usual methods, as perfect as possible; 
another, made by the new process, had required but 
twenty-two hours, and yet, while the microscope 
showed great irregularities in the former, none could 
be detected in this; the third was a similar screw 
before its final grinding. Professor Rogers produces 
a perfect screw by the following process: an ordi- 
nary, well-constructed lathe is used; and cuts of 
various depths are taken on a preliminary screw, for 
the purpose of tabulating the errors of the leading 
screw of the lathe as compared with a standard 
measuring-bar. This being done, a micrometer-screw 
is used to vary the relation between the leading 
screw and the cutting-tool. This screw is kept moy- 
ing automatically, or by hand, so as always to cor- 
respond with the tabulated values, which results in 
producing a screw nearly free from the errors of the 
leading screw. This screw is then ground with a 
nut cut in the same way; and, if not sufficiently 
perfect, it is then put in the place of the leading 
screw, and another screw cut from it by the same 
method, whereby any remaining errors are elimi- 
nated. A company has been formed for putting 
perfect screws on the market. 
In the animated discussion which followed, Presi- 
dent Sweet gave his experience in constructing the 
Cornell measuring-machine, and claimed that the 
nut should be made as long as the screw to avoid 
unequal wear of the latter. Among other opinions, 
it was claimed that scraping surfaces to a bearing is 
better than grinding; that tempered steel should be 
used, and other means devised for maintaining the 
screws perfect; and J. A. Brashear was referred to 
as having solved the problem of flat surfaces up to 
five inches diameter. 
Mr. W. E. Kent of New York presented rules for 
conducting boiler-tests, in which the precautions 
necessary for determining the actual heating-power 
of a fuel, or the efficiency of a steam-boiler, were set 
forth at length. A committee was appointed to re- 
port upon a uniform method of making such tests. 
Mr. W. B. LeVan resumed his advocacy of quick 
transit in a paper, ‘ New York to Chicago in seventeen 
hours,’ in which the time required for each of eight 
divisions was figured out, the average hourly mileage 
being fifty-five, whereas seventy to eighty miles is a 
common speed for short distances between Philadel- 
phia and New York. A change in locomotive valve- 
motions was also recommended. Mr. Charles E. 
Emery read ‘ Estimates for steam-users,’ in which he 
detailed the methods and formulae in use by his com- 
pany for arriving at the amount of steam furnished 
to various classes of customers. The New-York 
steam company has been selling steam at a fixed 
price since February, 1883. 
Mr. H. R. Towne, of the Yale & Towne company, 
explained their drawing-office system, by which all 
