4 Fishery Board for Scotland. 



place, and from one day or hour to another. While this is one of 

 the most fundamental and familiar of truths to the scientific student, 

 it is still worth while to lay stress upon it. For when the present 

 North Sea investigations were being begun, some twelve or thirteen 

 years ago, we were constantly told by certain persons that we could 

 never by any possibility come to learn the temperature, the salinity, 

 or other physical conditions of the North Sea ; for at the best our 

 observations could only be made at places miles apart, and repeated 

 at considerable intervals of time. It was obvious, we were told, that 

 of the temperatures in those miles of water between our chosen 

 stations out at sea, and of the temperatures at those stations on the 

 days when the observer was not there, we should know nothing 

 whatsoever ! Now, if this were so, there would be no such thing 

 as physical science ; but it is not so at all. There are, of course, 

 degrees of accuracy which are only to be attained by very close and 

 very frequent observations ; there are minute fluctuations of all 

 kinds which, if we desire to study them, we must do so by almost 

 microscopic methods. Were we studying, in the utmost possible 

 detail, the distribution of temperature on the land, we should be able 

 to note differences not only between the top and bottom of any 

 little hill, but • (as we all know) between one side of a garden and 

 the other. We may put all sorts of detail into a large-scale map 

 which we are unable to show upon a small-scale one. But the 

 minute features of a large-scale map, and the many local variations 

 which it illustrates, are only comparatively small and relatively 

 unimportant phenomena compared with those greater and more 

 general ones which underlie them all, and of which our small-scale 

 map gives us the broad and general picture. The principle of 

 continuity gives us the right to pass from one known point to 

 another, and from a few chosen points to draft the whole configuration 

 of a map. Even in the extraordinarily complicated phenomenon 

 which we call " Weather," where the secondary fiuctuations or minor 

 irregularities are exceptionally large, it is still possible to draw, day 

 by day, a " weather-chart " for a country or a continent, from a com- 

 paratively small number of observations, made at spots which are far 

 asunder. Through such observations, properly interspaced and not 

 too small in number, a principle of continuity is found to run, whicli 

 throws the minor fluctuations into the shade. 



When we have laid down in this fashion the mean distribution 

 of temperature over a sea or an ocean, the next step in our inquiry 

 is to study the seasonal changes of temperature in a similar way. Our 

 first chart represents the mean or average conditions of the year, 

 more or less approximately ; but we next want to know how June 

 differs from December, and any one month from another. To do 

 this we must, of course, multiply our observations, repeating them 

 year after year, for a term of years, and at all seasons of the year, 

 helped all the while by the same principle of continuity. For the 

 changes will be gradual and regular in Time, as we have found them 

 to be in Space, and so we eke out our knowledge, counting upon this 

 regularity or gradation. Even if at some station or other, for 

 example, we have no direct knowledge from observation of the 

 temperature during some particular month, but do possess such 

 knowledge for the months before and for the months after, we may 



