244 report- -1884. 



As the disturbance dies out the period of these irregular movements 

 increases, and waves with a period of two or three seconds have been 



recorded. 



For information respecting the velocity of propagation, I will refer 

 to my report of 1881, where some general results were given. ^ I shortly 

 expect to be able to give more definite information on this subject. 



Experiments on the Direction of Motion of a Point. 



Hitherto the only means that we have had at our disposal for de- 

 termining the direction of motion of a point, has been either to combine 

 the records of two rectangular components, or to trace a few of the more 

 conspicuous curves in a record given by a seismograph writing on a 

 stationary plate. Both of these methods can only be applied to prominent 

 vibrations in a record, and each of them, unless under special circum- 

 stances, is liable to error. The records given by seismographs with 

 single indices writing on moving plates, are for several reasons also open 

 to error, especially perhaps on account of the friction of the moving plate 

 exerting a drag on the recording index. To partially overcome these 

 difficulties, I have constructed a record receiver which works as follows : 



Shortly after the commencement of the disturbance, the smoked plate 



on which the index of a seismograph is writing, is suddenly^ dropped 

 vertically out of range of the index. It is next pushed along horizontally, 

 and then raised vertically back to its original level, so that it is again in 

 contact with the recording point. 



This operation is quickly repeated twelve times, at intervals of every 

 two seconds, so that twelve different diagrams are obtained on a strip of 

 smoked glass, each one being written on a different part of the plate. In 

 this way all effects of drag produced by the moving plate upon the pointer 

 are eliminated. As I have thus far only obtained one set of diagrams, I 

 must reserve a description of the results until a future occasion. 



The simultaneous observation of Earthquakes at three stations in 

 Telegrajihic connection. 



The advantages to ba gained by the observation of earthquakes at three 

 or more stations in telegraphic connection were first definitely pointed 

 out by Professor J. A. Ewing, in a communication to the Seismological 

 Society. A very similar method had, however, been previously followed 

 by Mr. T. Gray and myself, in our observations on artificially produced 

 disturbances. 



The method which I am now following is briefly as follows :— Near 

 to my house I have established, at the corners of a triangle, the sides of 

 which are each approximately 800 feet, similar instruments. These are 

 fixed on the heads of stakes level with the surface of the ground. The 

 records are written on smoked glass plates which at the time of an earth- 

 quake are drawn by means of a falling weight beneath the writing indices. 

 By means of electrical connections, these plates are simultaneously set in 

 motion by the withdrawal of a catch. As they move along, time intervals 

 are marked by levers deflected by electro-magnets every time a small 

 pendulum passes a cup of mercury. The pendulum, which is usually held 



