Notices respecting New Books. 1 4-7 



to tolerate. On the publication of the Catalogue of the British 

 Association, in which a systematic reformation of sidereal nomencla- 

 ture was carried out, it became a matter of consideration whether the 

 names used in the observed sequences should be altered to correspond 

 with that Catalogue ; but as it was found that this could not be 

 done without causing great confusion, the nomenclature of Bode i3 

 retained. Sir John observes that he regrets this the less, because 

 an entire remodelling of the whole system of constellations, both 

 northern and southern, is now necessary for this especial purpose — 

 without which the progress of this department of astronomy will of 

 necessity be slow and uncertain. 



The sequences or comparisons of apparent magnitude are given at 

 length. Their number is forty-six. They were observed at Feld- 

 hausen between July 16, 1835, and Feb. 25, 1838, and on board the 

 H.E.I. Company's ship Windsor, at sea, from March 28 to April 15, 

 1838. The number of stars included in each sequence is very 

 unequal, but may be stated as varying from twenty or thirty to 

 seventy or eighty. 



The sequences observed in the manner above described formed, as 

 it were, the raw material out of which a regular progression of mag- 

 nitudes was to be elaborated. The author remarks that " if obser- 

 vations of this nature could be made with perfect exactness, all 

 sequences in which the same stars occur should present them in the 

 same order of succession, in so far as it may not have been disturbed 

 by a real variability of the stars themselves. But as this is very far 

 indeed from being the case, owing to the many disturbing influences, 

 and the slight differences of magnitude between particular stars which 

 have to be discriminated, it is necessary to fix upon some criterion 

 by which the most probable order of arrangement can be discovered 

 among a number of discordant ones. Such a criterion is supplied 

 by an enumeration of cases." Thus if two stars, A and B, occur in 

 a certain number of sequences, and A is set down sometimes as su- 

 perior, sometimes as inferior to B, then the order of their arrange- 

 ment will be determined by placing the one first which has been so 

 placed the greatest number of times in the observed sequences. 

 Again, if there be three stars, A, B, C, and the direct comparisons 

 of B and C leave it doubtful which of those two is superior, and it be 

 found on a revision of all the observations that A has been set down 

 as decidedly superior to C but not to B, then there will be reason 

 for placing B before C. It is needless to go into further details ; the 

 general method of proceeding will be obvious. 



It might be supposed, both from the uncertainty which in a great 

 number of cases must attend the judgement formed from a naked 

 eye comparison, and the numerous discrepancies which occur in the 

 recorded observations, that the final result must be open to very con- 

 siderable doubt. This, however, does not seem to be by any means 

 the case ; on the contrary, to one who has not been in the habit of 

 making such comparisons, the precision will appear to be really 

 astonishing. " It might appear that so much is left to judgement 

 in such cases, as to induce uncertainty in the final results. But, in 



L2 



