CLYDE SEA AREA. 145 



terminating at the bar which separates Loch Goil, and at the entrance to Upper Loch 

 Long, just beyond the depression at Dog Roek. 



The set of observations in this division was usually an interrupted one, and rarely, if 

 ever, were all the stations studied on the same day. Consequently, the sections which 

 were drawn are somewhat less trustworthy than those for the lochs, and it has not been 

 considered necessary to reproduce them. They were, however, quite serviceable for 

 calculating the average temperature of the water on the occasion of each trip, and this 

 was done in order to compare the temperature changes as a whole with those of the other 

 divisions. Table LIX. gives the calculated temperature of the layers and of the- mass 

 of water as a whole ; and Table LX. shows the rate of change of temperature between 

 the dates of successive sections. The curves at the various stations did not differ much 

 from those for water of the same depth in the Arran Basin, except in the case of the 

 Dog Eock observations, which showed some affinity to those in the deep lochs. (See 

 Loch Goil, p. 122.) 



The time-depth diagrams for the station off the Gantock Beacon opposite Dunoon, 

 and for the Dog Bock soundings reproduced in figs. 10 and 11, Plate VI., show a general 

 homothermic change in the deeper layers with a slightly greater restriction at Dog Rock. 

 At the latter station contorted curves, showing layers of water varying irregularly in 

 temperature, were very frequently found, due probably to the outflow from Loch Goil 

 and Loch Long entering at different levels the mass of water in the Dunoon Basin. 



These diagrams show that in 1886 the temperature on the surface was above 50° from 

 June 28th to October 28th at Gantock, and from May 10th to October 25th at Dog 

 Rock, as compared with July 10th to November 6th in Loch Goil, where the greatest 

 depth reached by that temperature was 18 fathoms. In contrast, the bottom water at 

 Gantock was over 50° from August 27th to October 30th, and at Dog Rock from 

 August 30th to December 8th. The observations for 1887 and 1888 showed exactly the 

 same arrangement, with some slight differences of date. 



At Gantock the retardation of the date of maximum bottom temperature after 

 the surface averaged 27 days, at Dog Rock 40 days, and in Loch Goil, at Stuckbeg, 116 

 clays. The retardation of the minimum temperature at the bottom was 16 days after 

 that at the surface at Gantock, 37 at Dog Rock, and 38 at Stuckbeg. The contrast is 

 greatest between Dog Rock and Stuckbeg, the retardation of both maximum and mini- 

 mum being practically equal at the former, while at the latter the descent of low tem- 

 perature took place at the same rate, but the descent of the high temperature wave 

 took three times as long. The date of surface maximum and of surface minimum 

 corresponded closely for all three stations. The average number of days during which 

 heating and cooling continued for the two complete periods was in the order — Stuckbeg, 

 Dog Rock, Gantock— 186: 165, 169:168, 182:157 for the surface, and 255:105, 

 172 : 180, 189 : 171 for the bottom. Thus, at Stuckbeg, the water was 21 days 

 longer in heating than in cooling at the surface, and 150 days longer in heating than in 

 cooling at the bottom. At Dog Rock the time was the same, both for heating and 



VOL. XXXVIII. PART I. (NO. 1.) T 



