|Taught in 1999 by Jed Buchwald at the Massachusetts Institute of Technology
Replication of selected historic experiments: typical experiments might
be early electrometry; Colorimetry using Maxwell color circles; Fresnel,
Arago experiments on the interference of polarized light; Faraday and electromagnetic
induction; Hertz experiments on resonance between circuits. Reconstruction
of instrumentation and experimental apparatus based on close reading of
original sources. Emphasizes challenges of historiographic analysis, problems
of replication and theory verification as well as principles of transducers,
uncertainty analysis, data processing, and technical writing. Contrast
and comparison with laboratory practices and techniques of today. 12 units,
Institute Laboratory Requirement. J. Buchwald, L. Bucciarelli
This course has two related purposes. We want students to develop
an understanding of modern science's historical roots in a way that will
bring out the complex relationships between theory and experiment, and,
in particular, that will emphasize the hard work involved in the historical
establishment of scientific facts. To that end we will use a combination
of lectures with hands-on, laboratory work to bring out the skilled methods,
techniques and tacit knowledge involved in the building and conduct of
experiments, lessons that apply not only to historical material but also
to contemporary scientific work. We will explore the meaning of the distinction
between fact and theory in the sciences by linking our laboratory work
with the debates and claims made historically by the original discoverers,
asking such questions as whether, and in what sense, a fact is tightly
bound to a specific theoretical perspective, how anomalies arise and are
handled, and what sorts of conditions make historically for good data.
These questions, which have their seats in the particular historical experiments
that we will discuss and replicate, lead naturally to the second, related
purpose of our course. We want students to think through issues concerning
the sources of error, both systematic and random, that arise out of experimental
situations. Questions concerning the behavior of instruments, such as their
resolution, precision and accuracy, as well as the role and establishment
of standards will also arise naturally out of our investigations. We will
discuss questions of how laboratory reports are produced to persuade other
scientists that the investigator's claims are cogent, examining how the
character of laboratory reporting has changed over time. The difficult
and substantial issues raised by problems of replication will inform much
of our work - for example, whether a particular experiment is aimed historically
at replicating a novel effect, using substantially different apparatus
from that which had been used to produce the effect in the first place,
or whether the experiment is intended to be a close reproduction of the
15 to 18 students
three historical experiments, with 4 weeks devoted to each.
Each session will begin with two lectures presenting historical context
and relevant background information, for which the students will be given
a package of reading material containing original sources as well as relevant
Laboratory work begins during the second week. We will provide a reading
package containing copies of original documents, diagrams from original
sources, remarks concerning appropriate materials to use, and questions
to be addressed. We may, where appropriate, divide the students into two
or three groups, each of which will take a different point of view concerning
the issues in question. Laboratory work continues for two weeks.
In the last week of a given session, the students will bring their results
to class for discussion and analysis.
The list below is not complete, nor developed in particulars, but it does
represent historical experiments that we think will be instructive and
that can also be accomplished with reasonable facility. We do not intend
to provide the students with 'kits', but rather want them to produce the
relevant parts of the apparatus in much the same fashion that the original
experimenters would have to have done. Naturally this won't always be possible,
nor can we in all, or even many, instances use the same materials. We will
ourselves construct sample devices ahead of time and do trial runs.
a. Newton's early prism experiments: pass white light through prism at
minimum deviation, measure elongation of spectrum, pass segments of spectrum
through a second prism, determine refrangibility, use crossed prisms.
b. Bartholin and Huygens experiments on double refraction: use calcite
and early measurement techniques to explore differences from Snell's law.
c. Build a Wollaston refractometer to measure refractive indexes. Compare
measurement results with indexes determined using Brewster angle measurement.
d. Explore Malus' techniques for polarization: build early polarimeter,
examine with Biot et al. chromatic polarization.
e. Fresnel diffraction: reproduce Fresnel's early techniques, explore calculation
methods and limitations of early theory, move forward to integral methods
and high-accuracy observations (Fresnel reached .01mm, and I think we can
do this also). Could also do the "Poisson spot".
f. Fresnel and Arago experiments on the interference of polarized light.
g. Fraunhofer's "6 lamp" spectrum-mapping technique.
h. conical refraction: very interesting experimentally and mathematically,
crucial in early wave optics.
i. speed of light using beam-interruption methods (Foucault, later Michelson).
j. reflection of light from metals: not easy, mathematically a bit advanced,
but very significant for learning limits of experimental technique. Would
use an early polarimeter,
k. optical rotation in quartz: significant in early optics, highlights
phase issues, combines with experiments on phase-shifts in total internal
l. speed of light in a moving stream of water (Foucault).
m. Rayleigh scattering using a dust-filled chamber.
n. colorimetry using Maxwell color circles
2. Electricity, Magnetism, Electromagnetism, Electro-optics
a. mid-18th century Leyden phial experiments -- issues of charge transfer
b. Cavendish null-experiment on the inverse square. Great experiment, nice
issues of calculation, experimental limits, etc.
c. Coulomb torsion-balance electrometer. Vexing problems, good for learning
vagaries of devices. Combine with charge-density mapping of two spheres
placed near one another (later calculated by Poisson).
d. Faraday capacitor: measurement and discussion of specific inductive
capacity. Requires torsion-balance electrometer. Very significant on route
to field theory.
e. Voltaic pile in original form, including methods of testing for charge
f. Volta and electric quantity/intensity: origin of concept and techniques
for dealing with capacitance.
g. Ampere experiments on forces between wires connected to voltaic piles.
Magnificent null-experiments, not too hard to do, much learning potential
(including how to handle general bilinear force relations). Build an early
h. Arago's disk: later attributed to current induction by motion (disk
spinning under a suspended magnet drags magnet along). Raises many interesting
issues. Good recent literature.
i. Faraday and electromagnetic induction. Iron horsehoe, wound with wire,
use early ammeter.
j. Faraday and magnetic permeability: motions of diamagnetic and paramagnetic
objects in fields, raised important questions of 'polarity' vs. field explanations.
k. Faraday effect: rotation of plane of polarization of light passing through
a transparent dielectric (usually carbon disulphide). Lots of historical
resonance, not too hard to do.
l. Kerr effect: rotation of plane of polarization & production of ellipticity
in reflection of plane-polarized light from surface of a magnet. Very significant
historically, probably very hard to do. Uses a device that Zeeman later
used for Zeeman effect.
m. Hertz experiments on resonance between circuits: invention of resonator
and oscillator. Beautiful experiments, not too hard to do. Involves use
of induction coil.
n. Hertz's experiment to show the electrodynamic effect of rapidly-changing
dielectric polarization. Very nice experiment, requires rather large block
of pitch however. Probably could do it with asphalt, but must be sure the
substance used has very low conductivity.
o. Hertz propagation experiments as originally performed. Easy to set up
once resonator and oscillator in hand.
p. Hertz reflection and polarization of electric wave experiments.
q. Hall effect as originally performed. Would use different metals to probe
effect's character and possible interpretations.
r. Ohm's law using original devices and meanings.
3. Heat and Thermodynamics
a. Specific and latent heats using the Lavoisier-Laplace ice calorimeter.
b. Experiments on the law governing adiabatic gas expansion. Very instructive
since can be used to show among other things that a law now taken as a
canonical implication of energy conservation can be obtained on assumption
of heat conservation instead.
c. Joule-Thomson throttling experiment, showing no temperature change without
performance of work.
d. Thomson-Thomson experiment showing lowering of melting point of ice
with pressure. Important as an early implication of Carnot's claims.
4. Dynamics, Kinematics, Pressure
a. Conjectured Galileo inclined-plane experiment, using frets and a song
for timing to determine the law relating times to distances.
b. Chaldni figures for acoustics.
c. Bernoulli pressure experiments - variation of pressure in flow. Needs
some research since I am not certain that anyone actually did anything
like this in the 18th century, at least systematically.
d. The air-pump: construct one on 17th century 'schematics', examine various
situations, including the interesting one of "anaomalous suspension",
in which disks pressed hard together apparently do not separate under certain
circumstances when the chamber is evacuated. Used at the time as an argument
against Boyle's vacuum.
e. vis viva, vis mortua and the measure of 'force': relation between
depth of penetration of a falling object into a substratum like clay in
relation to height fallen.