INVESTIGATION OF TEMPERATUEE CHANGES. 255 



A^ery common, and perhaps natural, to generalize from insufficient data, or to 

 draw hasty inferences from isolated facts, and he gives several instances 

 illustrative of the way in which rash conclusions are or might have been 

 formed, and afterwards set at naught by later events.* 



The table of the years in which the great rivers of Central Europe Avere 

 frozen gives no assistance in coming to a positive conclusion as to any serious 

 change of the temperature in that region during the past 2,000 years. If 

 anything, the indications are tliat there has been an increased prevalence of 

 severer winters during the later centuries. Thus, for instance, the Seine is 

 recorded as having been frozen over at Paris, twice in the 9th century; not at 

 all in the 10th, 11th, or 12th ; once in the loth ; tirice in the 14th ; five times in 

 the loth; once in the 16th; five times in the 17th; fourteen times in the 18th; 

 and eleven times in the first half of the 19th. Similar facts are given in regard 

 to the Rhone. But it may justly be said that this apparent increase in the 

 number of times of freezing in later centuries may very possibly be due 

 to the fact that the records of such events kept in modern times are likely to 

 be much more complete, or, at all events, more accessible to research, than 

 those of earlier ages.f 



It may be thought that the facts stated in the preceding pages with refer- 

 ence to desiccation, especially in the region bordering on the Mediterranean, 

 indicate so rapid a change of climate, that the records of instrumental obser- 

 vations should — if the change were due in considerable part to diminution 

 of temperature — furnish some positive proof of such diminution. A due 

 consideration of all the facts will, however, lead us to the conclusion, that 

 over regions where the rain-fall has already become diminished to such an 

 extent as to be only just sufficient for the prosperous development of the 

 population, a small farther decrease will be attended by very serious conse- 



• For instance, in Provence the tlicnnometer between the years 1768 and 1788 never went lower than — 9° 

 (Cent.). This period of twenty years not liaving shown any such low temperatures as had been previously experi- 

 enced in that part of the country (of from —15° to — 18°), the idea became cuiTent tliat the winters were be- 

 coming milder; but in 1789 the thermometer sank to — 17°, thus indicating pretty clearly a return to former 

 conditions. 



+ A good illustration of the positive nianner in which statements about changes of climate are frequently 

 made, without any evidence being produced in their support, may be found in a recent charming work entitled 

 "Magyarland" (London, 1881), the author of which declares that the climate of Europe was "much colder" at 

 the time of the building of Trajan's bridge across the Danube [a. d. 10.3] than it now is. This assertion is accom- 

 panied by a statemenl, that that river was formerly frozen over habitually, so as to be >ised for the transport of 

 troops, while such an occurrence rarely, if ever, takes place at the present time. The tables in Arago's volume do 

 not sustain this assertion. 



