THE CONSTITUTION OF THE COPPfiR-TlN SEPJES OF ALLOYS. 37 
to make it evident that they once belonged to larger skeletons. Round these, and 
enveloping them symmetrically, is a complete system of fat rounded combs of 
which sometimes stretch right across the surface. Outside this /3 there is matter 
that we must suppose was not solid at the moment of chilling. The photograph does 
not show striation in the /3, but a somewhat deeper etch and a higher power brings 
out the striation perfectly. A photograph (fig. 25) at 45 diameters shows the 
striated ^ round the lobes of a, as well as the tin-rich material outside the material 
that was liquid at the moment of chilling. One can also see in one corner something 
which looks very like chill primary of The enormous growth of the a out of the 
solid y8 at lower temperatures can be best illustrated by the two following 
photographs, both at the same magnification. The first is a chill at 740'^ (fig. 26), 
that was not slowly cooled enough to dissolve quite all the a, a very little of which 
can still be seen in slender lines and rows of dots. It must be remembered that even 
this could have been dissolved by a slower passage through the C temperature. 
Contrast this with the other, a chill at 558°. In this (fig. 27) the enormous increase 
in the a and the jagged appearance of the crystals are well marked. There is no 
doubt that this growth of a took place in a rigid solid. 
It will be noticed that in the chill at 558° there is no residue of the striation effect 
so marked in the very slow-cooled chill at 775°. This is one reason why we think 
that the striation does not exist in the /3 until it has cooled below 500°. We think 
that as the chilling alloy cools rapidly through 500° a partial breaking up of the 
sometimes occurs and shows itself in the striation pattern. The two phases into 
which the /3 breaks up through the imperfection of the chill may not, however, differ 
so much in composition from each other as do the a and S of the slowly cooled Q' 
complex, for ignition which readily distinguishes between a and 8 does not readily 
bring out the striation even when we know it to be on a large scale. Another point 
is that the striation does not show itself in the chilled /3 of Sn 16 or 17, which has 
never been in contact with a nucleus of a. Under such circumstances we suppose 
that the state of supersaturation is more easily maintained. This view that the 
striation does not appear in the ji until it has cooled down to 500° is of some 
importance, as otherwise we are hardly entitled to speak of ^ as a uniform solid 
solution. Another argument is that now and then we find a chill containing 
unstriated y8, for example the patch of ^ of Sn 9 chilled at 546° and shown magnified 
280 diameters. This is quite free from striation, as are all the other patches in the 
section (fig. 18). 
The very symmetrical way in which the grows round the a (see fig. 24) during 
the a + liquid = /3 reaction, suggests the view that the two bodies may be 
isomorphous. Moreover, the primary combs of /8 seen in the chills of Sn 16 and Sn 17, 
taken between the solidus and liquidus, are indistinguishable in form from a combs, 
although they have never been in contact with that body. The facts already known 
with regard to solid solutions of saline mixtures would not have led us to expect such 
