]\gt ; [ 65 ] IV .
On II .
J. J. SYLVESTER , D.C.L. , F.R.S. , Professor of the of ( JAMrS )rMOND , Rcceived March April .
Continuation , to an nunnber of terrns , ) for " " This was sometime a adox , but now the time gives it proof scene IN the third section of this paper ( ' Phil. Trans A. , vol. 178 , p. 311 ) it was stated , on what is now seen to be insufficient evidence , that the asymptotic development of , the half of any Hypothenusal Number , could be expressed as a series of powers of ; the half of its antecedent , in which the indices followed the sequence .
It was there shown that , when quantities of an order of magnitude inferior to that of are neglected , ; but , on attempting to carry this development further , it was found that , though the next term came out , there was an infinite series of terms interposed between this one and , viz. , as proved in the present section , between and there lies an infinite series of terms whose indices are . . .
, and whose coefficients form a geometrical series of which the first term is and the common ratio We shall assume the law of the indices ( which , it may be remarked , is identical with that given in the introduction to this paper as originally printed in the ' Proceedings , ' but subsequently altered in the ' Transactions ' ) and write ESSOR .I .
.I .
SYLVESTER AND MR. J. HAMMOND \amp ; c. , ad ( 1 ) The of the coefficients will then be established by proving that If there were any terrms , of order superior to that of , whose indices did not obey the assumed law , any such term would make its presence felt in the course of the work ; for , in the , process we shall employ , the coefficient of each term has to be determined before that of any subsequent term can be found .
Tt was in way that the existence of terms between and was made manifest in the unsuccessful attempt to calculate the coefficient of .
It thus appears that the assumed law of the indices is the true one .
It will be remembered that , . .
, are the halves of the sharpened Halniltonian Numbers , . .
, and that consequently the relation ma be written in the form ( 2 ) The comparison of this value of with that given by ( 1 ) furnishes an equation which , alnter several reductions have been made , in which special attention must be paid to the order of the quantities under consideration , ultimately leads to the determination of the values of , . .
, in succession .
Taking unity to represent the order of , the orders of will be Hence , after expanding each of the binomials on the right-hand side of and arranging the terms in descending order , retaining only terms for which the order is superior to , we shall find the text above represents some unknown function , the asymptotic value of whose ratio to is not infinite .
ON HAMILTON'S A . . . . .
( 3 ) , , Again , retainiug only those terms of ) whose order is to , we have . . . .
( 4 ) Order 2 ; ; 1 ; ; From ( 3 ) and ( 4 ) we obtain by subtraction Order \mdash ; . . . . .
Changing , . .
, . .
respectively , equation ( 4 ) becomes so that , if we assume ) , the order of will be the as that of viz. , \mdash ; Hence , if we substitute for in ( 5 ) , neglecting in the ] quantities of the order , we shall find at , since order of does not exceed , we have anl in like so on .
equation ( 5 ) becomes A ( 6 ) where order ; ; ; Let them where terms as .
as , but not beyond , \mdash ; ( which is the order of ) have been retained .
Now con sists of terms whose orders are 2 , , 1 , . .
Thus the .
of is \mdash ; , and in the above expression all terms of superior to - have been retained , alud consequently ( rejecting the square of whose order is - in the line of ( 6 ) we may write ON HAMILTON'S NUMBERS .
In the second line of ( 6 ) we may reject the whole of ce its .
is \mdash ; , and write After .
values for the terms in ( 6 ) which contain , and at the same time dividing throughout by , we shall obtainl Order 1 \mdash ; , , . . . . .
( 7 ) We now write and where , observing that the values of and can be immediately deduced from those of and by changing , .
into , . .
, it is evident that and are both of the order - ; for and are both of the order \mdash ; .
Thus ( neglecting quantities whose order is equal to , or less than , ) we have ; \mdash ; \mdash ; ; ; Order ; ; ; Order ; and so Hence ( 7 ) becomes , . . . . .
( 8 ) Dividing this throughout by them Wl.iting and we obtain in exactly the same ) nerclS before , merely altering the letters in the previous work , Order ; ; where quantities of the order , or less , are now neglected .
Similarly Order ; and so on .
Thus ( 8 ) becomes ON Order 83 . . .
Now the erms of the highest order in this equation must vanish when we write and therefore \mdash ; , which gives .
Substituting this value for , we find Order \mdash ; \ldquo ; Et . . .
, which is a mere repetition of equation ( 8 ) , with all the letters moved forward one place .
Hence it is evident that , if we treat this equation as we treated ( 8 ) , we shall find , arriving , at the same time , at another equation which will be merely repetition of ( 8 ) , with all its letters moved forward two places ; and this process can be continued as long as we please .
Thus we arrive at the result\mdash ; and the asymptotic development for Hypothenusal Numbers is established .
Comparing this with the corresponding formula for Hamiltonian Numbers , given at the of the third section ( at the top of p. 302 , where the last term is incorrectly printed ) , it will be noticed that each of the two develol ) ments begins with an irregular portion consisting respectively of four and one terms , followed by a regular series .
In the one case the regular portion is , multiplied by a series whose general term is ; in the other it consists of a series of terms of the form multiplied by \mdash ;

