8 Temperature of Sea Affected bfj Diminished Transparency. 



perature which can never be subjected to calculation. We 

 can never know beforehand whether, in such and such a 

 year, these countless myriads of animalculae will be more or 

 less prolific, and what will be the direction of their migration 

 southwards. 



The phosphorescence of the sea is owing to minute animals 

 of the medusa kind. The phosphorescent regions occupy 

 very large spaces — sometimes in one latitude, sometimes in 

 another. Now, as the water of the phosphorescent spaces 

 is quite turbid, and as its diaphaneity is almost entirely 

 destroyed, it may become, by its abnormal heating, a cause 

 of notable disturbance in the temperature of the oceanic 

 and continental atmospheres. Who can foresee the intensity 

 of this cause of thermic variation ? who can ever know be- 

 forehand the place which it occupies ? 



Let us suppose the atmosphere immobile and perfectly 

 clear. Let us suppose, moreover, that the soil has every- 

 where, in an equal degree, absorbhig and emissive properties, 

 and the same capacity for heat; we should then observe 

 throughout the year, as the effect of solar action, a regular 

 and uninterrupted series of increasing temperatures, and a 

 corresponding series of decreasing temperatures. Each day 

 would have its invariable temperature. Under every deter- 

 mined parallel^ the days of the maximum and minimum of 

 heat would be respectively the same. 



This regular and hypothetical order is disturbed by the 

 mobility of the atmosphere ; by clouds more or less exten- 

 sive, and more or less permanent ; and by the diverse pro- 

 perties of the ground. Hence the elevations or depressions 

 of the normal heat of days, months, and years. As disturb- 

 ing causes do not act in the same way in every place, we 

 may expect to see the primitive figures differently modified ; 

 to find comparative inequalities of temperature where, from 

 the nature of things, the most perfect equality might have 

 been looked for. 



Nothing is better calculated to shew the extent of these 

 combined disturbing causes, than the comparison of mean 



