~ 112A6 ANNUAL REPORT. 
of different solubilities. We thus have the limits of temperature fixed, 
the lower one being that of the complete expulsion of carbonic acid and 
the higher one being that of the formation of hydraulic silicates and 
aluminates. 
It was not thought necessary, in view of the knowledge available, to 
undertake an extended series of experiments as to the temperature re- 
quired. But it was decided to heat a mixture of clay and carbonate of 
lime at different temperatures and to observe at what temperature the 
residue becomes constant. For this purpose a marl, from Harper, Ohio, 
was selected, whose composition is as follows: | 
Per cents. 
SiO ge ee oak eee 7a oc ec eae en 8.75 
ALOGERGION serach ES ee ak er arm ome Geet 2.78 
CACO ST IES Oe es OE Tee i ne ae ieee 79 .22 
MoCO gst Sit Oe ae OR ri ae fee Oats eee 2.67 
A sample of this material was heated to 1000°, 1050°, 1100° and 
150° C. with the results as shown by the following figures, the time 
chosen being 90 minutes, that is, the respective temperature was main- 
tained for this time. The amount of marl taken for analysis was 2 grams. 
Memperaturencaya. eee wale 1000 R O50 Ss eTO0r 1,150° 
Residue ncaitin cts seeres 2.06 1.00 OVA oR Bees 
The results were found to check sufficiently close in other burns. 
The samples were made up into briquettes. 
It is thus seen that practically all the marl was decomposed at 1100°. 
Although this series might have been extended to mixtures of clay and 
carbonate of lime, yet the fact that this marl, which is acknowledged to 
be a most excellent cement material, was practically decomposed, together 
with the data obtained from the work of Lunge and Schochor, Hertzfeld, 
Smith, seems sufficient to warrant the acceptance of this temperature as 
the standard. 
The question of the time of heating was one that next needed atten- 
tion. In order to determine the factor of time, the same marl was used 
and heated for 30, 60, 75, 90 and 120 minutes at the temperature of 
t100° C. with the following results: 
Times smINuUtgEshs cea cecum nyees hee 30 60 75 90 120 
Residuewsper cents. se cca ace 8.6) el 28, 2074 lO Ono 
In spite of the irregularity in the 90-minute burn, which shows an 
increase of 0.6 per,cent. of the residue, it was decided to adopt 75 
minutes as the standard length of the rational burn. The irregularity may 
be explained as one of the variations necessarily coincident with all 
rational analyses of this kind, which do not work upon finely pulverized, 
homogeneous samples. 
