150 
through the Royal Society Range is to a large degree 
answered by what I call the ‘ palimpsest theory” (v., 
pi" st7m): In effect the outlet glaciers flow down 
notches cut by earlier headward (or cwm) erosion. I 
hope to publish shortly a mass of evidence and illus- 
tration in support of this sequence in glacial erosion. 
GRIFFITH TAYLOR. 
Meteorological Bureau, Melbourne, July 26. 
Muret Sanders’s ‘‘Encyclopaidisches Wérterbuch”" 
gives “riegel,"’ in addition to the various ordinary 
meanings of the word ‘“‘bar,” including a bar of soap, 
eleven other meanings. What advantage is there in 
the use of a German term over an English term when 
both have equally varied meanings? The term 
“riegel’’ is especially overloaded, as in geography, 
according to Grimm’s ‘Deutsches Worterbuch,” it 
is used in South Germany for a ‘kleine Anhohe, 
steiler Absatz eines Berges,” and he also quotes its 
use for a watershed. 
Ordinary water erosion would certainly produce a 
slope with catenary curves if it is operating on suitable 
rock and under suitable conditions. 
The conclusion that the Discovery Hut was not 
erected as designed was not based only on Dr. Taylor’s 
photograph, and there could have been no difficulty in 
managing the supports on any surface of ice which 
had not so steep a slope as to be otherwise unsuitable. 
The more detailed information regarding the origin 
of the glacier valleys which Dr. Taylor obviously col- 
lected may, as was remarked in the review, explain 
their origin. Dr. Taylor’s further publication will be 
awaited with interest. Tue REVIEWER. 
ANNEALING GLASS. 
VERYONE who makes chemical apparatus 
~ by blowing glass practises annealing in a 
rude way by allowing the glass to cool slowly by 
g‘adual removal from the flame, or by the use of 
a smoky flame. In glass works more systematic 
annealing is effected by slow passage through a 
long chamber wherein the temperature falls from 
the incoming to the outgoing end. In the manu- 
facture of optical glass of many different qualities 
the question of annealing is one of the first 
importance, as they differ so much in fusibility. 
Messrs. Hilger have after a careful investigation 
found the means of arriving at the maximum 
temperature necessary, and also the necessary 
rate of cooling, which may progressively become 
more rapid. Optical glasses may differ as much 
as 200° C. in the maximum necessary tempera- 
ture, which temperature may be a long way below 
any visible softening point. It is desirable not 
to exceed the necessary temperature, as the very 
slow cooling at the higher temperature leads to 
great loss of time. 
The method adopted by Messrs. Hilger for test- 
ing different specimens of glass is interesting as 
an example of a physical investigation made with 
a view to practically useful results. The principle 
of the method can be described very shortly. 
Fig. 1 shows a bar of glass supported as a canti- 
lever, and carrying a load. Its edges are ground 
and polished in the form of two parallel planes. 
This is set up in an electrically heated muffle, with 
means for observing the temperature electrically. 
Polarised light broken up into interference bands 
NO. 2452, VOL. 98] 
NATURE 
[OcToBER 26, 1916 
| by passage through a Babinet’s compensator is 
passed through the glass, and when this is loaded 
the bands become inclined as shown in the figure, 
illustrating how perfectly the stress, whether 
of compression or extension, is proportional to 
the distance from the neutral axis. If the load is 
allowed to rest on a support in consequence of the 
slight yielding of the glass, the rate at which the 
bands change from the inclined to the straight 
position can be observed for any known tempera- 
Fic. 1. 
ture. Fig. 2 shows two specimens undergoing a 
change of temperature which sets up strains from 
the difference in temperature between the interior 
and the exterior. That the two specimens are very 
different is only too apparent. 
By watching the bands in specimens of glass 
Messrs. Hilger are able to ascertain when the 
glass is hot enough to allow the internal strains 
to be relieved in a convenient time, and whether 
as the glass cools internal strains are avoided by 
Fic. 2. 
sufficiently slow cooling. After a point is reached 
at which the glass has lost all viscosity the cool- - 
ing may be accelerated, and though the bands then 
become curved they straighten out again when 
‘ultimately the temperature is equalised. There 
is no hard-and-fast point at which the glass ceases 
to be viscous, and so there is a progressive per- 
missible increase in the rate of cooling. Messrs. 
Hilger have thus shown how annealing may be 
effected perfectly in the minimum of time. Though 
“~~ we 
