44 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. "JO 



peratures in the East Atlantic Ocean. An abnormally eastern posi- 

 tion on the other hand brings about low temperatures. If the abnor- 

 mal distribution of the air pressure is particularly strongly marked 

 it makes itself felt in the temperatures of the whole extent of the 

 ocean. Otherwise opposite temperature departures occur on the 

 opposite shores." 



Petersen comes to the important conclusion that the variations in 

 the Gulf Stream influence directly the Icelandic air pressure mini- 

 mum; but not so much — as Meinardus had assumed — in that they 

 themselves increase or diminish its intensity by means of systems of 

 self-inducing- force, but that they alter the situation of it. An inten- 

 sified Gulf Stream drift leads greater quantities of heat into the 

 Norwegian Ocean, warms the air, generates an air pressure mini- 

 mum and draws thereby the Icelandic minimum towards the east. 

 In this way more westerly and northwesterly winds are generated 

 in the Atlantic Ocean, which tend to hinder the Gulf Stream. If, 

 however, the Gulf Stream in consequence of these flows more 

 weakly, this has the opposite influence and the Icelandic minimum 

 has a tendency to retreat towards Greenland. As an accompany- 

 ing phenomenon it occurs that when the Gulf Stream is weakened, 

 then the cold east Greenland current, which is its compensation cur- 

 rent, flows slower and extends less far than commonly. In this 

 way the Icelandic air pressure minimum may more easily be pressed 

 back toward the west into a relatively warmer region and so the 

 cyclical process begins again. In this way it is that " the Gulf 

 Stream by means of its indwelling forces regulates its own trans- 

 portafion of heat and forms a current w^hich alternates between a 

 time of strong flow and a time of weak flow." 



Petersen thinks that his tables on the temperature anomalies at 

 his twelve stations in the twelve months of the year prove the impos- 

 sibility of the assumption that variations in the surface temperatures 

 can be produced " by variations in temperatures of tropical waters 

 which are carried along through the Atlantic water circulation 

 througho'ut its whole course with the velocity of the water flow." 

 One sees no such movement of negative or positive anomalies from 

 one station toward the other. He may be right in maintaining that 

 most variations are not thus caused, but he has not proved that they 

 can never be caused in this way. He ignores in his conclusion the 

 source of error that without proof he has assumed that the water 

 masses move from west towards east in the same direction as his 

 steamer lines. If on the contrary the current goes at right angles 



