XIV, 6 King: Philippine Bast-fiber Ropes 507 
tions of a millimeter, but only to the nearest whole millimeter, 
which was sufficiently accurate for the purposes of this inves- 
tigation. Each test specimen was measured close to each splice 
and in the middle, and the average of the three values used. 
Diameters are not usually given in rope-test data. Direct 
diameter measurements would be very much more uncertain 
than girth measurements, especially if taken by inexperienced 
workmen, on account of the judgment required. To obtain 
such measurements numerous trials with the caliper must be 
made, and special care exercised to secure two diametrically 
opposite points, both in the boundary line of the strand and 
in the circumference of the circumscribing circle. The so-called 
“diameter” of the test specimens given in this paper was not 
measured directly and is not the true diameter, but was cal- 
culated upon the assumption that the girth measurement of the 
rope is a true circle. The relation of the girth to the true cir- 
cumference and the relation of the so-called “diameter” to the 
actual diameter are discussed in a former paper.’ 
Actually, the diameter so calculated is a little less than the 
diameter of the circumscribing circle, but the difference is slight 
and the value is probably more accurate than could be obtained 
by direct measurement. 
Cross-sectional areas.—Even with the most carefully machine- 
laid cordage, cross-sectional areas and values based on them are 
ambiguous unless accompanied by adequate explanations, and 
such data are very infrequently given in practice. The pecu- 
liarities of two- and three-strand cordage, shown in text fig. 1, 
more marked in the bast rope tested than in abaca or other 
filamentous fiber ropes, that make difficult the measurements of 
the actual solid, sectional area, are (1) the irregular trefoil 
Shape of the rope section and (2) the appreciable void space 
in the strands themselves. The actual area contained within 
the three separate strands is much less than the area of a 
circumscribing circle, and the relation between the two is dis- 
cussed in a previous paper.’ There is an approximately constant 
ratio between these two factors, as is shown in Table I. In 
all of the tables in this paper the actual measurements of average 
girth were made in millimeters, of weight per meter in kilograms, 
and of breaking load in pounds, and the equivalents recorded 
are in round numbers, accurate to a fraction of a per cent. 
* King, A. E. W., Philip. Journ. Sci. § A 13. (1918) 285-839. 
* King, loc. cit. 
