800 PROF. A. H. GIBSON ON THE LOSS OF ENERGY AT OBLIQUE IMPACT 
velocities is the same as occurs at a sudden change of velocity between the same limits 
in a circular pipe suffering an enlargement of section. In view of the very different 
conditions obtaining in the two cases, there would appear to be as little theoretical as 
experimental justification for this assumption, and the present experimental investigation 
was projected with a view of determining the magnitude of the loss and its variation 
with the angle of impact, with the velocities of the streams, and with their relative 
cross-sectional areas. 
2. DESCRIPTION OF APPARATUS AND MerHop oF EXPERIMENTING. 
The pipes used in the investigation were rectangular in section, being made of wood 
and varnished. The following table shows the range of sizes and of angles examined :— 
Sections of 
Series. Ratio of Areas. Values of 6. 
Impinging Stream. | Primary and Final Streams. 
1p }” wide x 1” deep 3” wide x 1” deep eel 5°, 15°, 30°, 45°, 60°, 90°. 
1 fear eo WN DP ET SG racy Dial 15°, 30°, 45°, 60°, 90°. 
TE a ee anes ee ae cle. 5 onal 15°, 30°, 45°, 60°, 90°. 
IV. BE 2 SE ag DP SAE ee 4:1 45°. 
Nis a eget sen Die gs OS Uae Ged 60°, 90°. 
Fig. 1 shows a sectional plan of one of the channels. Pressures were measured at 
points (1), (2), and (3), by water piezometers at low velocities, and by differential water 
and mercury piezometers at high velocities. Point (3) is distant 12 inches down-stream, 
and points (1) and (2) each 1°5 inches upstream from the junction. The piezometer 
openings in the walls of the passages were formed as shown in fig. 1, a The pressure 
