340 Scientific Proceedings, Royal Dublin Society. 



of the ordinates of these and running a curve through the points 

 thus found. It is symmetrical with regard to the equinoctial line, 

 and represents equally well the changes in the length of the day 

 in the two halves of the year, and on comparison it will be found 

 to coincide exactly with the "dawn" curve on the clock-dial. 



I therefore conclude that the curves were made in or about the 

 latitude 34° 6*5', and that they represent the variable hours in the 

 manner described above. 



The Japanese call the six periods into which their days and 

 nights are divided " Tokis," and each toki is again divided into 

 two. They number them from midnight and noon in a somewhat 

 curious manner. The hours are struck on a bell or drum, but as 

 the numbers one, two, and three strokes are used only as signals 

 for the army and in convents, they have to do without these in 

 their clocks, and they get out of the difficulty in the following 

 way : — Midnight is the first " Toki," accordingly they subtract 1 

 from 10 and strike 9. For the second " Toki " they subtract 2 

 from 10 and strike 8, and so on till noon, when the same process 

 is begun again. 



The figures on the right-hand side of the dial represent these 

 numbers, as shown on the two dials I have drawn. 



The second clock, though slightly different in construction, 

 depends on the same principle. At one side of the slit through 

 which the pointer projects is a series of 13 indices, with the same 

 Japanese symbols engraved upon them which have already been de- 

 scribed in the case of the first clock. These indices are capable of 

 moving up and down, so as to suit the varying length of the hours 

 at different dates. At the other side of the central slit is a rever- 

 sible dial, on one side of which is a scale with intervals corresponding 

 to the hour spaces on the 5 th line from the left in the clock already 

 described. This side of the dial carries symbols at the top of it 

 denoting " 2nd half of the 4th and 6th months." On the other 

 side of the dial the intervals on the scale correspond to the hour 

 spaces down the line of dots between the 4th and 5th vertical hues 

 on the first dial, the symbols at the top representing the words, 

 " 1st half of the fourth and seventh months." From this it is 

 clear that the clock was intended to have seven of such reversib]e 

 dials, which would cover the whole 24 periods into which the year 

 is divided. The scales on the two sides of this dial supply two 



