60 
MESSRS. C. T. HEYCOCK AND F. H. NEVILLE ON 
of the tin-rich material; but the chill primary, when examined with a higher 
power, is a mass of rounded bars, and has these inclusions in it. We thus see that 
this chill primary was formed at lower temperatures than the large combs, and did 
not altogether escape the transformation of the G temperature. 
Sn 27. S.c. chill at 675° (not reproduced). 
This is similar to the preceding except that the large combs occupy a greater area 
and there is less of the chill primary. We do not give photographs of these upper 
chills, as the y is sufficiently well seen in succeeding alloys. 
Sn 27. S.c. chill at 647° (fig. 69). 
In this the combs of large primary have almost entirely filled the ingot before the 
moment of chilling. It is quite possible, however, to detect here and there the gaps 
where three lobes of primary meet, and to see in such places a trace of chill primary 
which is now completely bar-shaped. The surface, when strongly etched, assumes 
the patchy appearance of a chemically uniform mass of grains, but in the photograph 
which we give this effect is barely visible. 
Sn 27. S.c. chill at 635° (fig. 70). 
This ingot proves that Sn 27 is capable of existing as a uniform solid solution of 
y without mother liquid. The G transformation has not yet taken place, the state 
of the alloy at the moment of chilling being represented by a point, very near y, in 
the angle between the lines cf and £(/. A few dark spots represent drojDlets of 
tin-rich mother-substance still liquid at the moment of chilling, and these contain 
chill primary consisting of rj bars; this liquid is not sufficient in amount to invalidate 
the above statement that the ingot is a uniform solid solution. 
Sn 27. Chill at 625° (fig. 71). 
Here we see the same alloy as that of the preceding figure but chilled ten degrees 
lower, after, mstead of heforc, the G transformation. The uniform y has broken up 
into bars (really plates seen edgewise) of 17 , and at the same time liquid, represented 
in the figure by dark lines and rows of spots, has been formed. 
As Professor PvOOZEBOOM has pointed out to us, Ave can calcidate approximately the 
amount of the G liquid formed in any alloy l)y the reaction jf — CugSn + liq^. For, 
assuming that / is at Sn 27 and G at Sn 42, let us consider the changes in Sn 26. 
As the temperature falls through the short interval lietween the transformation curve 
c/ and the fC line, the following gradual change occurs :— 
Ciq^Sno 0 = 0’5 CuygSiqj fi- 0’5 Gu^gSii^y, 
