1875.| Physics. 125 
attained; and on cooling, the after-glow caused a momentary and violent de- 
pression of the liquid index (Fig. 5). A crepitation in the wire was heard at 
the moment this change took place. Moreover, the critical temperature was 
Fie. 5. 
approximately that at which iron lost its magnetic character upon heating or 
regained it on cooling. Further Prof. Tait has shown that at this temperature 
the thermo-eletric condition of iron changes, and that at a lower temperature 
nickel undergoes a similar change. Hence it is most probable that nickel and 
cobalt will be found subje& to analogous molecular changes at the temperature 
wherein their transition from the magnetic state takes place,—nickel at a 
lower temperature than iron, and cobalt at a higher. Experiments have also 
been made to exclude the effe@ of oxidation during the heating of the wires. 
Iron and steel wires have been enclosed in atmospheres of hydrogen, 
nitrogen, and carbonic acid gas, and in each case similar effects are produced 
to those already described. In carbonic acid the gas is decomposed around 
the heated wire, and brilliant scales, apparently of graphite, are deposited on 
the iron. The foregoing remarks the author begged to be regarded as merely 
fragmentary observations, preliminary to a more detailed investigation which 
he hoped shortly to undertake. After Prof. Barrett’s paper had been read the 
following discussion took place. 
Prof. CLERK MAXWELL said—In theoretical investigations an assumption is 
often made that if you make a slight change in the circumstances a slight 
change will be produced in the result; but this is not always the case in 
Nature. It seems that in the cooling of iron wire, described by Mr. Barrett, 
something goes off at a certain moment—something tumbles over at the 
moment of the kick. Other things tumbled over too,—traps for catching mice, 
for instance,—and the parts might not have the power of getting up again. 
If a row of bricks were placed on end, and the first one knocked down, it 
would knock down the next, and so the whole row would tumble down. 
When I was a boy we used to call this “Sending Jack for mustard.” This 
mode of the transference of motion was not the same thing as wave motion, 
because in a wave the particles swing back, but the bricks have to get up 
again how they can, and won’t return to their former position automatically. 
So in Prof. Barrett’s experiment, a critical point is reached where a break 
down occurs among the molecules. In the process of cooling iron wire 
something happens which makes a noise, makes the wire lengthen, and make 
it glow again. A slight change in the circumstances has made a great change 
in the phenomena. Seeing Mr. Herbert Spencer present reminds me he has 
written that something of the kind may take place among the nerves, though 
I do not agree with the assumption he has made. 
