868 
correct for glancing incidence. Although the presaure on CD is 
possibly close to v, as assumed in (2) and (3), it seems to be 
difficult to estimate it more accurately for the gauges in current 
use. 
5. A second neglect in (2) and (3) is the possibility of "ir- 
regular or Mach reflection" (see (7)). Since so-called regular 
reflection in water ceases at an angle of incidence near 65° for 
ehockse of 300 atmospheres (8), Mach reflection from side-on gauges 
is also to be expected in the pressure range which interests us. 
In fact, the complete exvlanation of the doubling nersdox referred 
to in the previous paragreph may depend upon non-linear theory, 
even for very weak waves (see von Neumann (7), page 8). 
§. ‘In order to avoid the unsolved theoretical problems associ- 
ated with the side-on orientation, the cese of normal incidence 
4s considered here. In this position reflection is regular and 
the diffraction problem is much easier. There is also experi- 
mental, as well as theoretical, reaeon for preferring the head-on 
orientation, since diepersion is less there than for the side-on 
position. According to reference (9), page 20, the standard 
deviation is from 5 to 30% more in the side-on orientation. The 
increased dispersion is presumably due to the same causes: ir- 
regular reflection and diffraction. 
7. The ultimate problem is to infer the préssure-time curve 
of a shock from the deformations of several copper pellets. To 
do this it is necessary to know the response of the pistone res- 
vonsible for these deformations. This response is determined by 
ee Sites 
