310 SURVEY OP THE COAST 



hoops, making the ribs of the sphere, and covered with 

 white linen. Their diameters were from sixteen to twenty 

 inches. They were very distinctly visihie with the telescope 

 on my instrument, t!ioiigh only thirteen inches in length, at 

 a distance of lifty miles, aiid as far as ten miles with the na- 

 ked eye ; but they did not answer equally well in the lower 

 atmosphere of this coiintiy near the sea shore, as was to 

 be expected. 1 was therefore induced to use a kind of sig- 

 nals presenting aluminous point by the reflection of the sun, 

 and adapted to the situation of that body at such times of the 

 day as from the general state of illumination appeared to be 

 most favoura!)le to observation. In the middle of the day, 

 the illumination will not serve at all for distinct vision, and 

 even the largest objects become indistinct, on account of the 

 vapour, (as this is con monly called) ; or rather because the 

 reflection of the light from all objects goes upwards, and does 

 not meet the eye of the observer on the surface of the earth. 



As the cheapest reflecting surface which I could choose 

 was sheet tin, and the construction of si)lieres became more 

 difficult and expensive ; and the spheres themselves always 

 presented a small point, I chose the form of a truncated cone, 

 under such an angle as would be the most favourable for the 

 morning and evening illumination. 



The next point wliich appeared desirable with respect to 

 the signals was to find their places in case they shoidd be 

 removed. The means which 1 considered as best adapted 

 for the purpose were the following : — 



Plate VI. fig. 4. — aa.bb is a truncated cone of tin, the 

 height ab equal to nineteen inches, the lower diameter aa 

 equal to seventeen inches, the upper diatneter bb equal to 

 fourteen inches; the top e is a horizontal tin plate of three 

 inches diameter, elevated five inches above the dianjetrr bb 

 which serves to nail the signal to the top of the pole ef, and 

 from which to the diameter bb it forms a truncated cone of 

 a greater vertical angle. The pole ef. upon which the signal 

 is fastened, is about three or four inches in diameter, and of 



