The MeduscG of Millepora. 



3 



that the coincidence is not accidental, but represents the separation of 

 three distinct types of wave motion having different rates of propa- 

 gation. 



On plotting the records it is found that the time curves of the first 

 two phases form curved lines, indicating an increase of apparent 

 velocity with distance from the origin, such that, applying Eudzki's 

 investigation, the wave motion represented by these two phases must 

 have travelled through the earth, along curved wave paths, convex 

 towards the centre of the earth, and with a rate of propagation which 

 increases with the distance from the surface. On continuing these 

 curves, by extrapolation, to the origin they give rates of propagation 

 in very fair concordance with the rates of propagation of condensa- 

 tional and clistortional plane waves which' may be expected to obtain 

 in continuous rock at some distance from the surface of the earth. 



The waves of the third phase show no such increase of rate of 

 progagation with distance from the origin. The rate of propagation is 

 uniform at all distances ; from which it is concluded that the great 

 undulations of the third phase are surface waves, travelling with a 

 uniform rate of propagation round the surface of the earth. It is also 

 found that the waves of this phase set up by great earthquakes travel 

 faster than those set up by lesser ones, and from this it is concluded 

 that the rate of propagation of these waves is in some way a function 

 of their size, thus affording a confirmation of Lord Kelvin's suggestion 

 that their propagation is in part gravitational. 



The general conclusion is that in the complete record of a distant 

 earthquake, three distinct types of wave motion can be recognised (1) 

 condensational, and (2) distortional plane waves, travellihg by brachi 

 stochronic paths through the earth, and (3) elastic, or gravitational 

 elastic, surface waves, travelling round the surface of the earth. The 

 records are, however, often incomplete by the omission of the first or 

 the first and second of these phases, and the widely divergent estimates 

 of the apparent rate of propagation of the preliminary tremors 

 are largely due to this. 



The Medusae of Millepora." By Sydney J. Hickson. F.RS. 

 Eeceived November 7, — Eead November 23, 1899. 



Since the discovery of male medusa in specimens of Millepora 

 collected by Professor Haddon in Torres Straits eight years ago (2), I 

 have examined several large collections of specimens, both dried and 

 preserved in spirit, from different parts of the world with the object of 

 comparing the medusae and the ampullae they form in the varying 

 forms which the genus exhibits. The examination of the dried coralla 

 in museums has convinced me that the presence or absence of ampullae 



B 2 



