Fish and Drought 431 



the sense that the bottom was exposed to the air, but n< 

 theless moist enough to be called wet. Owing, probabh 



:^ht general rise of the ground-water of the neighbourhood, 

 enough water had been able to filter through the mud of the 

 bottom and to rise to the surface and overflow, producing a 

 very shallow pool (q), not more than two or three centimetres 

 deep or more than a metre across. The wetting of the mud 

 below by this infiltration must have aroused the sleepers, who 

 then all started .to rise at the same time. But when they 

 released themselves from the mud there was not enough water 

 to float them all, and a formidable struggle for existence 

 going on at the time of my visit, and the quantity of dead and 

 <l\in^ li>h lying all round the edges of the pool furnished 

 sufficient evidence of its fierceness. The poor fish would no 

 doubt have willingly re-buried themselves and so saved their 

 when they perceived their mistake, but the stronger ones, 

 which were in possession of the only part of the pool whuh 

 could be called liquid, kept shouldering them outwards on to 

 the mud, where they died in the air. When I Irft tin- struggle 

 was still going on, and it looked as if the level of the water was 

 falling, so that it i^ unlikely that many, if ,tn\ . would be able 

 to retrieve their mistake by burying themselves again. As 

 I left for England the next morning I was not able to continue 

 my observations. 



Although the years vary much as regards humidity, and in 



Mipply <>f water in the ditrh ha- often fallen 



to a pretty low ebb, I was informed that tin- l.t^t time that the 



ditrh became D the year 1814, almost a hundred 



years ago; therefore the experience of the summer of i, 



n a new one for all the fish in tin- ditch. \ 



manoeuvre of protective self -burial was carried out without a 

 ler to accomplish this a fine in-tn.. i was 

 necessary to perceive the impending drying up of the c. 

 and then to , omi he operation betimes so as to t-.m-h it 



before desicc M eomplcte. 



It inu-t be reme: area of canal having the 



muddy bottom, whi< h alone is capable of receiving the fish. 

 is very restricted ; and fr..m the number of theM- that . am< 



