206 
The Philippine Journal of Science 
1921 
nor the favorable clean and rough bonding surface characteristic 
of freshly crushed Sisiman stone and screenings. In both of 
these series of tests, the weakening effect of salt-water immer- 
sion is unmistakably evident. Seventy-seven cubes made of con- 
crete employed in the United States Army Quartermaster Pier 
are still submerged in Manila Bay, and eighty-eight specimens are 
being kept in air under the eaves of the materials testing labor- 
atory for long-time tests. Data on these test pieces will be pub- 
lished as soon as possible. 
Some other mixtures used in Manila and tested at the Bureau 
of Science are the 1 : 2.5 : 5, the 1:3:6, and the 1 : 1.5 : 4.5. 
The last proportion was used largely in constructing the Masonic 
Temple, and the results obtained from field-made test specimens 
are very poor. The compressive strength of test pieces made 
of 1:1.5: 4.5 concrete and aged 28 days averages as low as 
254 pounds per square inch and never exceeds 673, the grand 
average being 446. With the material available much better 
results should have been obtained. Older test pieces in this 
series show practically no gain in strength, the grand average 
obtained from specimens aged 36 to 49 days being only 483 
pounds per square inch. Both of the leaner mixtures 1 : 2.5 : 5 
and 1:3:6 gave better average results than did the 1 : 1.5 : 4.5 
mixtures. At ages ranging from 28 to 34 days the 1 : 2.5 : 5 
mixtures gave an average compressive strength of 729 pounds 
per square inch, which at 38 to 41 days is augmented to 1,005. 
The 1:3:6 results average 739 pounds per square inch at 28 
to 36 days; this strength is practically the same as that given 
by the 1 : 2.5 : 5 specimens. One cannot help but be impressed 
here, as in other instances, with the sameness in the average 
compressive strength given by specimens made of concrete that 
is apparently differently proportioned. For all practical pur- 
poses there is no difference between the average strengths of 
the 1 : 2.5 : 5 and 1:3:6 concretes ; and richer mixtures, made 
of the same aggregate and the same cement, which should 
show higher strengths, on the contrary, give abnormally lower 
values. These irregular results seem to point to a faulty method 
of measuring the cement, so that many of the mixtures tested 
are deficient in this material. Practically all of the sand used 
in Manila for concrete work comes from the Pasig and Mari- 
quina River beds. Pasig River sand is fine and is composed of 
very soft grains of rock in advanced stages of decomposition. 
It contains little if any quartz and is often contaminated with 
considerable shell debris. It is a poor concrete aggregate and 
