715 



llie riglitliaiid end, JO to tlie gieeii. The time is Middle-Java lime. Tiic 

 snns-heigiit and azimuth were calculated and with these the readings 

 of the theodolite were reduced. Az„ and iiw stand for the observed 

 azimuth and height of the arc; A„ is the distance of the observed 

 points from the sun calculated from the 4 foregoing columns ; A„. is 

 the angle between the suns vertical and the radiusvector from the 

 sun to the arc, deduced from tiie observations. 



The column under A„ shows, that the points measured deviate 

 sensibly from the ring, for the red of the ring is formed at a 

 distance of 45°6' from the sun. The mean deviation is 1.1°. Gradually 

 the distance decreases, but for Nr. 12 it is still 0.6° larger tiian thai 

 of the ring. This deviation is so big and so systematic, that it is 

 impossible to think of observational errors. Indeed there is question 

 here of a no7i tangent arc. Tiie position of the tangent-point of the 

 arc was calculated according to Bkavais's theory. The results lia\e 

 been entered under hb, Ab and Ab- The calculation was carried 

 through for the 10"' observation for green (?z = 1 .31 1 5) for the rest 

 for red (»= 1.307). In taking the differences between observation 

 and calculation the tirst four points, which in consequence of tlie 

 initial weakness of the arc happened to be less accurate than the 

 others, were combined to a mean value. The observations 5 and 6, 

 8 and 9, which refer to the ends of the arc, were substituted by 

 their mean values. 



Almost all the observed points are too high (column h,v-b gives 

 the difference observation and calculation), but they approach the 

 height calculated from theory. The angle A, which according to 

 theory should increase for a sinking sun, in reality rapidly decreases. 

 In consequence the difference between observation and calculation 

 decreases from 10° lot 1°. Finally, the distance from the sun remains 

 almost constantly 0.9° too big, hardly showing any tendency to 

 decrease. 



Durinii the inhole time 0/ observation the arc remains outside of 

 Bravais's arc ; the position with respect to the snn approaches more 

 anil more that of the theoretical tangent point. 



This arc deviates from that of Bravais and hence still more fronr 

 that of Pernteh. No more is it in harmony with Exner's theory. For in 

 this case we have to assume a normal plane inclined at an angle 

 of 30°. In our case the rays of the sun are in their turn inclined 

 to this plane at an angle of at least 57.1°— 30° = 27.1°. The 

 smallest distance from the arc to tiie sun is then 57.6°, which is 

 quite out of question for the observed arc. 



47» 



