screws of the reversing frames needed but 2 or 3 turns to release their hold. Probably in the 

 case here referred to the frame reversed at a depth of from 30 to 50 metres. 



After this occurrence, and after the above explanation had been found for it, piano 

 wire was substituted, when both frame and thermometer at once did their work properly. But 

 now another difficulty arose, no less serious than the former, which also had its origin in the 

 construction of the reversing frame, and which was not found out, until the frame had been 

 used several times, and after the loss of a lead and a thermometer. 



A sounding was made with a Negretti and Zambra frame. The object being merely to 

 ascertain whether the depth averaged 2000 metres, and the wire broke when the thermometer 

 had been brought up to within 15 metres of the surface. A careful examination of the broken 

 end of the wire showed that it was twisted off by a lefthanded revolution. 



The explanation of this accident can only be found in the fact that the screw of the 

 reversing frame, being released when the hauling up commences, is afterwards locked, and 

 continues so during the remainder of the ascent with the result that frame, wire, and lead 

 assume a lefthand revolving motion, clue to the reaction of the propeller on the frame. 



At first the twists are divided over the entire length of the wire. But when the lead 

 has come nearer to the surface, and the speed of heaving in the wire is gradually reduced for 

 the sake of safety, the twists in the wire, which has now become quite short, increase rapidly, 

 as the lead continues revolving, presuming that the weights have not been dropped ; the wire 

 eventually becoming twisted to such an extent as to prove fatal to the weakest point in the 

 line, i. e. that portion of it which is fastened and hielden in the splice, and hence cannot be 

 kept under regular control. Here it was that the breakage occurred. 



A careful consideration of the causes of these two accidents led to the following conclusions. 



1 . \\ T hen usincr a' twisted wire-line for reversino- thermometers on the Neeretti and 

 Zambra system a righthanded screw should have a righthand-twisted line, and a lefthanded screw 

 a lefthand-twisted line. This matter should be attended to when the instrument is constructed. 

 If circumstances do not permit of this it is urgently necessary that the speed of descent should 

 be and remain great from the moment that the thermometer touches the water. Any hitch in 

 the working of the sounding machine might cause the untimely reversing of the thermometer. 

 At all events it is clesirable that the screw should not be able to release its hold too quickly, 

 certainly not till after several turns, and over a considerable distance. This can cause no 

 appreciable difference in the registration of temperatures at great depths. 



The reversing frame might also be improved, by an arrangement which would insure 

 the screw's acting only with an upward motion regardless of revolutions due to the twistings 

 of the line. This might be done by introducing a small horizontal hinged-plate, which when 

 the frame is at rest would act as a catch on one of the screw blades, and which could only 

 be forced out of position by the pressure of the water when the frame has an upward motion. 



2. It is a great mistake to lock the screw in the reversing frame .after it has been 

 unscrewed in the hauling up ; the screw should be able to continue to turn loosely, in the same 

 manner as is done for instance in the waterbottle with thermometer, system Pettersson. 



