. MUSEUM OF COMPARATIVE ZOOLOGY. 171 



Having, therefore, tested the value of the improvements, it only re- 

 mained for me to put them in a better mechanical shape, and I therefore 

 submit the accompanying drawings, which' have been withheld for the 

 past two years that any slight idea occurring to me might be embraced 

 in the general plan. 



Description of the Parts of the Machine. 



The Reel {A, Figs. I., 11., III., IV., V.) should be, for convenience, one 

 fathom in circumference of drum, and should have a friction score which 

 is V-shaped in cross section. When the sinker strikes the bottom, the 

 momentum of the reel and its remaining' wire requires to be quickly 

 overcome by the resistance of the friction line, in order that any undue 

 slacking and consequent kinking of the wire may be avoided. To secure 

 this quick stoppage, the reel should be made as light as the very con- 

 siderable strength required for reeling in will permit. In reeling in 

 without the rceling-in or strain pulley, it is necessary that the reel 

 should be of sufficient strength to resist the accumulated strain of suc- 

 cessive convolutions of the wire under strong tension. If a reel can be 

 made of light weight and yet capable of bearing this crushing force when 

 reeling in from great depths, the strain pulley for reeling in will no longer 

 be needed. The reel should be made of iron or steel, as a brass reel 

 enwrapped with steel wire would perhaps cause the deterioration of the 

 wire through the effects of galvanic action. It is rigidly attached to its 

 axle by a key, and the easy removal of the key would be a convenience, 

 since the reel without its axle could be stowed into a much smaller 

 tank than is necessary when the axle is retained. When a reel contain- 

 ing wire is out of use it is generally kept in a tank of oil. A crank 

 should be provided for each end of the axle. 



The Register (B, Fig. I.). — The construction of this is sufficiently ex- 

 plained by the drawing. It is the same as that used by Sir William 

 Thomson, and is worked by a screw-thread attached to the axle of the 

 reel. The register evidently does not record fathoms, an interpolation 

 being necessary in determining, from the reading of the register, the 

 length of wire payed out. It is very handy, however, in keeping an 

 approximate account of the wire payed oflf the reel, so that the correct 

 amount of resistance can be applied to the reel. The correction to 

 add to the reading of the register is so easily found after the manner 

 followed on board the " Blake," that the register is desirable also as a 

 check on the odometer. 



The Reeling-in or Strain Rxdley (C, D, E, Figs. I., IV., V.) is com- 



