is, 2 King: Philippine Concrete and its Aggregates 215 
suits are 760 and 927 pounds per square inch, respectively, and 
the mean of the entire series is 831 pounds per square inch. 
These specimens range in age from 33 to 36 days. Two Zam- 
boanga sands were tested at the Bureau of Science. Both are 
coarse, but they show strikingly different mortar-strength values. 
The beach sand composed of fairly hard grains yields a 1 : 3 
mortar which at 28 days gives an ultimate compressive strength 
that is 90 per cent of that shown by the Ottawa sand test spec- 
imens. The river sand, containing soft grains in an advanced 
state of decomposition, gives a 1 : 3 mortar that is weak and at 
28 days is only 46 per cent as strong as Ottawa sand mortar. 
GENERAL DISCUSSION 
Attention has already been called to the extraordinary vari- 
ation in ultimate compressive strength of field concrete specimens 
made of the same materials, nominally proportioned alike, and 
differing only in being mixed and cast on different dates. Differ- 
ences amounting to several hundred per cent characterize the 
results obtained from specimens from nearly all the provinces. 
With uniform materials, carefully proportioned and thoroughly 
mixed, such variation would be much reduced. Exceedingly 
crude methods of proportioning probably account for the extreme 
variations noted. 
Considerable variation also exists in the compressive strengths 
obtained from specimens made from a single batch of concrete, 
as well as from specimens made of the same materials, propor- 
tioned alike, and made from various batches of concrete mixed 
on the same day. It will be noticed that most of the ultimate 
compressive strength values appearing in Table 8 are averages of 
two or more results obtained by rupturing two or more test 
specimens. It is the difference between results that have been 
so averaged that is being discussed at present. One would 
expect more or less close agreement of results obtained from 
specimens made from the same batch of concrete. However, the 
variation from the mean compressive strength in a given series 
in the field specimens is sometimes abnormally high, and in one 
case actually amounts to 135 per cent. Such extreme variation 
from a mean value in a series indicates either improper mixing 
or careless proportioning of materials; slovenly procedure dur- 
ing the molding of the test piece may also be a contributing 
cause. Any one, or two, or all three of these factors could cause 
erratic results. In Table 9 are recorded the maximum and 
minimum compressive strength values of concrete specimens in 
