308 



Prof. B. Stewart and W. Dodgson. 



[Nov. 20, 



our base station. We have in all 27 inequalities between — 7"0 and 

 + 6*0, the sums of which represent the numbers of the vertical 

 column headed Kew in Table III. In like manner for Toronto we 

 have 27 similar inequalities. Let us now add together algebraically 

 the inequalities of Kew and the corresponding inequalities of Toronto, 

 under the supposition that the phases are simultaneous at the two 

 places. We thus get a series of 27 inequalities representing the 

 united result of the two observatories. Treating these in the same 

 way in which single observatories are treated in Table III, we 

 get 27 sums, and now finally let us add these sums together. We 

 get in this particular instance, Kew + Toronto= 133657. Let us 

 next, on the supposition that a phase at Toronto occurs 6 days earlier 

 than the corresponding one at Kew, rectify this by pushing each 

 Toronto inequality 6 divisions to the right before adding L it to the 

 corresponding Kew inequality. Obtaining as before 27 sums and 

 adding them together, we get as a result 141794. 



This then is greater than the result obtained by adding the in- 

 equalities together as they stand, and would seem to indicate that by 

 means of thus pushing Toronto to the right we are bringing the phases 

 more nearly into accordance with one another. 



7. In the following table we have exhibited the results obtained by 

 this method of treatment. 



Table IV. — Algebraic sum of Kew and Toronto Temperature 

 Inequalities. 



Kew and Toronto (together) =133657 



,, (Toronto pushed 6 divisions to right) =141794 



7 „ =143880 



8 „ =144746 



9 „ =142523 

 10 „ =137556 



It thus appears that we get the greatest sum, and consequently the 

 nearest approach to similarity of phase when we push Toronto for- 

 ward, or to the right, 8 divisions — in other words similar phases take 

 place at Toronto about 8 days before they take place at Kew. 



Viewing Kew as our standard, we have in like manner compared 

 Utrecht and Kew together. And just as we might expect Toronto to 

 be before Kew, so we might expect Utrecht to be behind it. It will 

 be seen by the following table that this surmise is correct. 



