588 KEPORT — 1888. 



6. On the Temperature of some Scottish Rivers. 

 By Hugh Robert Mill, D.Sc, F.B.S.E. 



Observations were carried on in a number of rivers in different parts of Scot- 

 land, with the result that from January to July 1888 the variations of temperature 

 were practically identical ; all the curves showing a series of well-marked maxima 

 and minima exactly coincident in date and approximately equal in amplitude, 

 although the actual temperature varied for each stream. The maxima appear to 

 characterise periods of high air-temperature and slight rainfall over Scotland gene- 

 rally, the minima periods of low air-temperature and considerable precipitation. 

 The temperature of the water in the Aray, the one stream of the west coast exam- 

 ined, was always lower than that of the air ; but over the east coast rivers the 

 air-temperature was more extreme, and frequently fell below that of the water. 



7. On the recent Magnetic Survey of Japan. 

 By Professor Caegill G. Knott, D.Sc, F.B.S.E. 



This survey was carried out during the summer months of 1887, the expenses 

 of the expedition being wholly borne by the Imperial University. 



The object was to obtain, within as short a time as possible, a general survey 

 of the whole of Japan. The work was accomplished by two parties, known as 

 the Northern and Southern Parties. The Northern party was under my own 

 charge, and its route may be roughly described as a cycle of that part of Japan 

 which lies to the east and north of a line drawn from Fujiyama north-west to the 

 peninsula of Noto. Mr. Tanakadate took charge of the Southern party, whose 

 stations all lay to the west of the line just described. The Northern party took 

 observations at four stations in the Island of Ezo, and the Southern party made a 

 trip to Korea and obtained a series of valuable measurements in the neighbourhood 

 of Pusan. 



The Northern party were equipped with a complete set of instruments of the 

 usual Kew pattern — the dip circle used being one which was kindly loaned by the 

 Kew Committee of the Royal Society. The chronometer was checked daily by 

 sextant observations. The Southern party used Mr. Tanakadate's own form of 

 electro-magnetic declinometer (described * Proc. R. S. E.' 1884-6), which was also 

 fitted up for the measurement of the horizontal force by the ordinary method of 

 sines. The theodolite, whose base was an essential part of Mr. Tanakadate's 

 decUnometer, served in its usual form for taking the transits and altitudes necessary 

 for rating the chronometer and finding the true meridian. 



Generally speaking, the magnetic features of Japan present great irregularities, 

 a fact which the highly volcanic condition of the country would lead us to expect. 

 The south-western portion of the main island, vrith the adjacent islands fringing 

 the Inland Sea, presents fairly uniform magnetic features. The regions where the 

 greatest disturbances exist are (1) the great central mountain region to the north 

 and north-west of Fujiyama, and (2) the region included between the 38th and 

 40th parallels of latitude. 



The general characteristics of the iso-magnetic lines corresponding to the obser- 

 vations made are as follows : — The lines of equal dip, of equal horizontal force, and 

 of equal total force are approximately straight, while the lines of equal declination 

 are distinctly parabolic or hyperbolic, approximating very closely to the general 

 form of the main island. Out of the eighty-one stations at which complete observa- 

 tions were carried out, a selection of fifty was made ; and the values of the magnetic 

 elements at these selected stations were combined by the method of least squares 

 according to the usual mode. Linear expressions in latitude and longitude were 

 assumed for the dip, horizontal force, and total force ; and, for the particular case 

 of the declination, a term was added in the square of the longitude. The latitude 

 and longitude co-ordinates were referred to the mean station (36° 30' N. lat. ; 

 137° 9' E. long.). 



The mean formulse so obtained are as follows : — ^ and X being the latitude and 



