670 EEPORT— 1882. 



better effect could be obtained by the use of Argand burners with chimneys, than 

 from any burner of the ordinary flat flame type. The paper concluded with some 

 remarks on ventilation, and on the way in which heat generated by the combustion 

 of gas coidd be utilised for ventilating purposes. 



6. On Sound Signals. By E. Price Edwakds. 



The purposes for which such signals are employed are chiefly in connection with 

 railway and sea travelling. No recent improvement as regards railway sound 

 signals can be recorded, but there has been a great development of late years with 

 marine sound signals. Railway whistles are very distressing to the public ; not so 

 to the callous engine-drivers who use them without any regard to the public com- 

 fort. This is a matter which might be ventilated by the press, for there is a great 

 deal of unnecessary whistling, and there is no necessitj^ for the use of the piercingly 

 shrill sounds which are commonly employed in this country. The explosive signals 

 used in foggy weather are admittedly inefficient and costly, besides Ijeing annoying 

 to railway travellers. For signals between ships at sea, the recent alteration of the 

 regulations for preventing collisions has necessitated the use of sound signals to a con- 

 siderable extent, and various instruments operated by manual labour have been devised 

 to meet the requirements as regards sailing vessels ; steamers of course use the steam 

 whistle. For coast fog-signals there has been a great progress made of late years ; 

 whistles are not used for this purpose, reed horns have been discarded, and now 

 sirens and explosives only are used. A novel feature of late with the siren is the 

 introduction of two notes as a means of distinction. Each signal is to have a com- 

 bination of high and low notes so that it may proclaim its own individuality and 

 be readily recognised. It is also proposed to make double explosive signal rockets, 

 by means of which a further distinction may be made for this class of sound 

 siprnals. 



7. Some of the Causes of Collision at Sea. By Captain Colojib, B.N. 



The causes assigned in the Annual Wreck Register, according to the official 

 statement for the year 1879-80, were divided into two classes : — (1) conditions 

 which are irremediable and inevitable ; and (2) the moral obliquities of seamen. 

 The former comprise 40 per cent., and the latter tlie remaining 60 per cent. 



The classification is unsatisfactory because collisions can rarely be set down to a 

 single cause ; and because, though inevitable conditions are always present, they do 

 not necessarily produce ine-\-itable accidents. 



There has never yet been any public enquiry into the causes of collision, nor 

 any official statement of the conditions under which they happen, further than the 

 state of the atmosphere and sea, and the question of whether darkness or daylight 

 prevailed. Contrary to what might have been supposed, a thick atmosphere does 

 not increase, but generally markedly decreases the number of collisions ; while, as 

 might be supposed, the greater number of accidents, and those which are most dis- 

 astrous, happen at night. It is convenient to confine the attention to these latter. 



It has been found by examining carefully considerable numbers of collision cases 

 which have been tried in the courts, that in only about 7 per cent, of them were 

 the ships first seen at a greater angle to the direction of the courses steered than 

 45°; and in only about 26 per cent, of the cases did the two ships' courses cross at 

 a greater angle than 45°. The normal condition previous to collision is when the 

 courses of the two approaching ships are nearly, but not quite, opposite. Collisions 

 under any other conditions are rare. The conditions of sea traffic make it quite cer- 

 tain that one or both ships approaching imder these conditions turn towards each 

 other. But such a movement or movements very rarely indeed precede collision. 

 On the other hand, the contrary movement of one or both ships turning away from 

 the other is nearly always found preceding collision. 



Constantly improving experiments have been made to measure exactly in time 

 and space the movements of ships under the action of their rudders. It is found 



