SURFACE AND OTHER EFFECTS 393 



is taken up in section 10.2. An interesting observation is the occur- 

 rence of later domes caused by secondary or bubble pulses, if the charge 

 is fired at sufficient depth for the gas bubble to contract one or more 

 times before venting. These secondary domes are much smaller and 

 less regular than the shock wave dome, because of the smaller pressure 

 and their superposition on earlier disturbances, which obscure them. 



For shallow charges, the products of explosion retain their identity 

 in the gas bubble until this bubble reaches the surface and venting oc- 

 curs. The size and state of motion of the bubble at this time both 

 change greatly with the initial depth, and as a consequence the effects 

 of venting also vary markedly with depth. The result of venting of the 

 bubble gives rise to plumes of spray mixed with explosion products, two 

 distinct types of plume being formed under different conditions: 



(1). A relatively narrow vertical plume of variable height and 

 velocity, depending on the charge depth. 



(2). Radial plumes projected outward in ah directions through the 



spray dome, the number and development of which depend on the 



depth of the charge. These plumes, if present, appear at the same 



time as or later than the vertical plume. 



The differences in plume formation with depth are clearly shown in 



the sequence photographs in Plate XI for 300-pound TNT charges fired 



at depths of 18, 26, 40, and 65 feet. These photographs are reproduced 



for the same intervals of time after firing, but the distance scale is not 



the same for the different sequences. The first phenomenon of the 



spray dome appears at all depths, its height and duration decreasing 



with depth as is to be expected. At a charge depth of 18 feet, the gas 



bubble for this size of charge reaches the surface before its contraction 



begins and has a small upward velocity of migration. The venting 



therefore is nearly radial and the radial plumes of Plate XI result. At 



a depth of about 26 feet, however, the bubble reaches the surface when 



smallest and moving upward most rapidly, and almost all the water 



above it is thrown up vertically to form the narrow high plume of Plate 



XL 



At greater depths the vertical plume becomes increasingly less de- 

 veloped and the radial plumes reappear. This change in plume forma- 

 tion is simply understood as a result of the fact that venting of the 

 bubble occurs at progressively later stages of its second oscillation; the 

 vertical motion becoming smaller and the bubble motion more nearly 

 radial as its size increases. This development reaches its maximum at 

 a depth of about 40 feet, where the plumes are nearly all radial, as shown 

 in Plate XL At greater depths the vertical plume reappears and 

 reaches maximum development at a depth of 65 feet, corresponding to 

 venting at the second contraction of the bubble with maximum upward 



