Fig. 1 Philos. Trans.396 .
B Fig. 2 .
I. A short Account of Negativo-affirmative Arithmetic , by Mr. John Colson , F. R. S. THE Usefulness of this Arithmetic consists in this , that it performs all the Operations with more Ease and Expedition than the common Affirmative Arithmetic , especially in large Numbers : And it differs from the common Arithmetic chiefly in this , that it admits of Negative Figures promiscuously with the affirmative .
These negative Figures are distinguished from the Affirmative , by the Sign placed over them .
---Thus 3709286573961472 is one of these Numbers , which may be converted into its Equivalent common Number 2308726432039468 , in this manner : 3009006503000470 0700280070961002 2308726432039468 ( 1 .
) Write down all the affirmative Figures by themselves , putting a Cipher in the place of every negative Figure .
( 2 .
) Write down all the negative Figures by themselves , putting a Cipher in the place of every affirmative Figure .
( 3 .
) Subtract the last Number from the first , and the Remainder will be a common Number , equivalent to the given negativoaffirmative Number .
See the Operation above .
But the readiest practical Way of performing this Reduction in any given Number , will be in this manner : Begin at the left hand ; and going over all the Figures in order , observe these Rules .
( 1 .
) An afY firmative firmative Figure before a Negative must be diminished by an Unit .
( 2 .
) A negative Figure before an Affirmative , must be changed into its Complement to 10 .
( 3 .
) A negative Figure before a Negative , must be changed into its Complement to 9 .
All other Figures must remain unchanged , and a Cipher is always to be understood where there is no significant Figure .
The sign of the Cipher is neglected ; but where there is occasion to consider it , it is always supposed the same as the Sign of the following Figure .
Thus the negati--vo-affirmative Number 729586455982001730 is immediately reduced to 710585545977998330 ; and so of all others .
But on the contrary , common Numbers may be reduced to negativo-affirmative Numbers a great variety of Ways , by substituting instead of the Figures 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , their respective Values 19 , 18 , 17 , 16 , 15 , 14 , 13 , 12 , 11 , in any places at pleasure .
But the most useful Reduction of this kind is what I call A Reduction to small Figures , which consists in throwing out all the large Figures , 9 , 8 , 7 , 6 , out of any given Number , and introducing in their room the equivalent small Figures 11 , 12 , 13 , 14 , respectively .
Thus 182937462 may be reduced to 223143542 , consisting only of small Figures .
But this Reduction may be performed more readily by these Rules following .
( 1 .
) A small Figure before a large Figure must be increased by an Unit .
( 2 .
) A large Figure before a large Figure must be changed into its negative Complement to 9 .
( 3 .
) A large Figure before a small Figure must be changed into its negative Complement to 10 .
Other Figures are not to be changed ; and 5 will be ambiguous , being to be esteemed either large or small , according as the Figure following is either large or small .
Some Examples of this Reduction shall here follow , both in whole Numbers and Decimal Fractions .
--37068259764 = 43132340244 -729528960739957 = 1331 5310 41340043 --9260872395,87294 = 1134 1132404,133 14 --Or ( 9 ) 926087239587294 = ( 10 ) 11 34113240413 314 ---(m ) 387916407953 , &amp;c. = ( m ) 412124412153 , &amp;c. It is to be observed , that in this last Example the Numbers are what I call interminate , or Approximations only ; that is , the first and most valuable Figures are expressed , and all the rest ( whether finite or infinite in Number , whether known or unknown ) are omitted as inconsiderable , and insinuated by the Mark &amp;c. Also the Index m before the Number stands for some Integer , expressing the Distance of the first Figure 3 or 4 from the Place of Units ; which Integer is either affirmative or negative , according as the said first Figure stands in integral or fractional Places .
The Example immediately before is a particular Instance of this .
And thus much by way of Notation : To proceed therefore to the Operations to be performed with these Numbers , whether reduced to small Figures or not ; and first of Addition .
Place the Numbers to be added just under one ano there , observing the Homogeneity of Places , as in common Numbers .
Then beginning at the Right Hand , collect the Figures in the first Row or Column , acoording to their Signs , and place the Result underY 2 neath : neath : And so successively of all the other Columns , as in Example 1 .
But if at any time this result cannot be expressed by a single Figure , it may be writ down with two or more Figures , observing the Homogeneity of Places , and then the Sum may be collected over again .
But to save this trouble , it will be sufficient to reserve the Figure in mind which belongs to the next Column , and collect it with the Figures of that Column ; as in Examples 2 , 3 .
If the Numbers to be added are reduced to small Figures , as in Example 3 , their Addition will be very simple , and the Sum may also be exhibited in small Figures , by an easy Substitution of Equivalents , where there is occasion .
( 1 .
) ( 2 .
) ----257384263 647039682 -----709821370 49 8273651 --5807305 819403765 --------953 364212 18646 43894 ( 3 .
) ---(m ) 215 314043 1213 , &amp;c. ---(m ) 50420 31425512 , &amp;c. ----(m-1 ) 431023102413 , &amp;c. -(m-2 ) 51342110321 , &amp;c. ( m-3 ) 2130421032 , &amp;c. -(m-4 ) 132021224 , &amp;c. -(m-5 ) 13224315 , &amp;c. ---(m+1 ) 13333214134312 , &amp;c. Subtraction in this Arithmetic is reduced to Addition , by changing all the Signs of the Number to be subtracted .
-Thus , if from ( n ) 72938429637 , &amp;c. we are to -subtract ( n-2 ) 810735926 , &amp;c. the Remainder will be found as in Example 4 .
( 4 .
) -(n ) 72938429637 , &amp;c. ( n-2 ) 810735926 , &amp;c. --(n ) 737471 54343 , &amp;c. Thus in all Cases will Addition and Subtraction be easily performed : But the chief use of this Method will be , to ease the trouble of prolix Multiplications .
And here , as well as in Division , the first and most valuable Figures may be first sound , and consequently the Product may be continued to as many Places as shall be required , without finding any unnecessary Figures ; which is a convenience not to be had in the ordinary way of Multiplication .
Let it be proposed to Multiply together the Numbers 8605729398715 and 389175836438 , which reduced ---to small Figures will be 11414331401315 and 4112 -24244442 .
Write down these two Numbers one under the other upon a slip of Paper , with the Figures at equal distances , and then cut them asunder .
Take either of the Numbers for a Multiplier , and place it over the other in an inverted position , so as its first Figure may be just over the first Figure of the Multiplicand .
Moveable Multiplier 411224244442 ---Multiplicand -11414331401315 ----4561791825498645062606080 -11113 1124 2632 1 2311 ------Product = 4650861937096017072623170 Then Multiply these two first Figures together , and their Product ( 4 ~1= 4 ) place underneath .
Then move your Multiplier a place forwarder , so that two of its first Figures may be over two of the first Figures of the Multiplicand ; and collecting their two Products ( 4 ~1+1~1= 5 ) put their Result underneath in the next place .
Move the Multiplier a place forwarder ; and collecting the three Products ( 4 ~4+1~1+1~1= 1 6 ) put the Result underneath , as in the Example .
Move the Multiplier , and collect the four Products ( 4 ~1+1~4+1~1+2~1= 11 ) which write underneath as before .
And so proceed by one stop at a time , as long as any Figures of the Multiplier can be over any Figures of the Multiplicand .
Lastly collect the Product into one Line , which being reduced to a common Number will be 33491419369039969 27377170 .
From this Process it may be observed , that at every new Situation of the moveable Multiplier those Figures only are to be Multiplied together , each by each , as are found over one another .
And this Multipli tiplication is to be performed , and the several Products collected , according to the Rules of Specious Multiplication , wherein like Signs will make + , and unlike Signs will make in the Product .
This will always make the Products destroy one another , or at least will depress and keep them low , and the Figures themselves being always small , the Result will be always small , and often but a single Figure , which is the great Compendium of this Method .
When an Approximation only is desired , or when the Product is to be produced to a given Number of places , the Operation may be continued one place farther , in order to obtain so many Places true as are required .
For seldom any Correction extends beyond the place immediately aforegoing , and that is generally corrected but by an unite , and very often needs no Correction at all ; which will be of no small convenience in the Multiplication of Decimal Fractions .
In this Method we may ( if we please ) begin the Process of Multiplication from the lowest places , or from the right hand , as is usual in common Arithmetic , and then the Correction may be carried on continually to the next place , and so the Product may be always comprehended in one Line , without the use of any Superfluous Figures .
Of this I shall give an instance in the foregoing Example .
Moveable Multiplier 411224244442 Multiplicand ---11414331401315 3349141936903996927377170 = Product Place the moveable Multiplier inverted in such a manner , as that its last Figure 2 may be just over 5 the last Figure of the Multiplicand .
Multiply these together ( 2 ~5= 10 ) and set down the last Figure of the Product 0 just under , reserving the first Figure 1 for the next place .
Then move the Multiplier a place forwarder , so that two of its last Figures may be over two of the last Figures of the Multiplicand , and then Multiplying and collecting you will have 1+2~1+4~5= 17 .
Set down 7 in the next place of the Product , and reserve 1 .
At the next remove you will have 1+2~3+4~1+4~5= 31 .
Set down 1 and carry 3 .
Then 3+2~1+4~3+4~1+4~5 -= 23 = 37 .
Set down 7 and carry 3 .
Then 3+2~ 0+4~1+4~3+4~1+4~5=3= 17 .
Set down 7 and carry 1 .
And so proceed as long as there can be any Figures over one another , and the Product will be found as before .
This way of Multiplication is so easy , and may be made so samiliar by a little Practice ; that it will be but little short of Multiplication by Inspection , and will doubtless seem very surprising to those who are only acquainted with the common tedious way of Multiplication : especially if we content our selves with a mental preparation of the Numbers given , or only mark those Figures that are to be changed , which by some Practice is easily attained .
The first of these two ways of Multiplication will be most convenient for interminate Numbers .
As if we were to multiply ( m ) 307149741748 , &amp;c. by ( n ) 183609712649 , &amp;c. the Product will be found ( m + n ) 563956758222 , &amp;c. as may appear from the Process following .
~ ( m ) 313150342352 , &amp;c. / ( m + n ) 644057258378 , &amp;c. Here the Index of the first Figure of the Product will be m+ n , or the Sum of the Indexes of the given Numbers ; but it would have been m+n+1 if there had been any increase from the Product of the two first Figures , or if there had been any correction to have been made to the Cipher , which is understood before the first Figure of the Product .
When both the Numbers to be multiplied are interminate , as in the last Example , they ought to consist of the same number of Places , or otherwise the greater number must be reduced to the lesser , by cutting off the superfluous Places : And the Product is not to be continued beyond the same number of Places .
If but one of the number is interminate , the other must be reduced to the form of an interminate Number , either by cutting of the excess of places if it has more , or by supplying or supposing Ciphers , if it has fewer places than the interminate Number .
Then the same restrictions will take place as before .
The Method of Division in this Arithmetic will not be so simple or expeditious as Multiplication .
After a trial of several ways , I think this following will be the most commodious .
Reduce the Dividend and Divisor to small Figures , and form a Tarifsa or Table of all the Multiples of the Divisor as far as 5 .
Compare these Multiples with the Dividend , and with the several Remainders after the Multiples have been Subtracted , by which means you will discover the several small Figures and their Signs , to be put successively in the Quotient .
To form the Table of Multiples , set down the Divisor above , drawing a line , under which set down the Divisor over again , putting 1 over against it .
Add these two together according to the Rule for Addition in small Figures , and put 2 over against their Sum .
Add this last and the Divisor together , and put 3 over against their Sum .
Then add this last and the Divisor together , and put 4 over against their Sum .
Lastly , add this and the Divisor together , putting 5 over against their Sum .
Thus will you have a Table of all the Multiples of the Divisor , as far as will be necessary .
Thus for Example , if ( m + n ) 563956758222 , &amp;c. = ( m +n+ 1 ) 1444043242222 , &amp;c. is to be divided by ( n ) 183609712649 , &amp;c. = ( n ) 2244 10313451 , &amp;c. the Process will be as here follows , by which the Quotient will be sound ( m ) 313150342354 , &amp;c. = ( m ) 307149741746 , &amp;c. TABLE of Multiples .
( n ) 224410313451 , &amp;c. 1 ( n ) 224410313451 , &amp;c. 2 ( n ) 433221425302 , &amp;c. 3 ( n ) 551231142153 , &amp;c. 4 ( n + 1 ) 1334441251404 , &amp;c. 5 ( n + 1 ) 1122051443245 , &amp;c. Quotient = ( m ) 313150342354 , &amp;c. Dividend = ( m +n+ 1 ) 1444043242222 , &amp;c. 551231142153 , &amp;c. 13132420335 , &amp;c. 22441031345 , &amp;c. 15233451010 , &amp;c. 5512311421 , &amp;c. 335140411 , &amp;c. 224410313 , &amp;c. 111331324 , &amp;c. 112205144 , &amp;c. 1534220 , &amp;c. 551231 , &amp;c. 124451 , &amp;c. 133444 , &amp;c. 3205 , &amp;c. 3332 , &amp;c. 1533 , &amp;c. 551 , &amp;c. 124 , &amp;c. 112 , &amp;c. 12 , &amp;c. The Index of the Quotient is found by subtracting the Index of the Divisor from the Index of the Dividend , when the first Figure or Figures of the Divisor are not greater than the like Figures of the Dividend .
Thus ( m + n ) 56 , &amp;c. / ( n ) 18 , &amp;c. = ( m ) 30 , &amp;c. When they are greater , then an Unit must be farther Subtracted from the Index of the Dividend .
Thus ( m +n+ 1)144 , &amp;c. / ( n ) 224 , &amp;c. = ( m ) 313 , &amp;c. Now a little to illustrate this Process , it may be observed , that the Dividend 144 , &amp;c. or 56 , &amp;c. being compared with the several Multiples of the Divisor in the Table , it is easily perceived that 55 , &amp;c. makes the nearest approach to it , and therefore its correspending Figure 3 must be made the first Figure of the Quotient , which I place just over .
Under this I place its respective Multiple , changing the Signs and collecting , because it ought to be subtracted .
Then the remain der 131 , &amp;c. being compared with the Table , I find the nearest to it ( whether in excess or defect ) to be 22 , &amp;c. or 18 , &amp;c. belonging to 1 .
This therefore is made the second Figure of the Quotient , and its Multiple with the Signs changed is placed under , and collected with the last Remainder .
The Result , or new Remainder 152 , &amp;c. that is 152 , &amp;c. or 52 , &amp;c. being compared with the Table , the nearest Multiple is 55 , &amp;c. belonging to 3 , which 3 therefore is made the next Figure of the Quotient , but with the Sign over it , because this Remainder is Negative .
And for this reason the Multiple 55 , &amp;c. is put down in its place without changing its Signs .
The rest of the Process will be obvious enough , and if any scruple arises about placing the Numbers , it may easily be removed moved by a little attention to their respective Indexes .
In the Division of interminate Numbers , the same restrictions are to obtain , as are already mentioned in Multiplication .
And this may suffice for a short Summary of Nega tivo-affirmative Arithmetic , as to the ordinary Ope rations of Addition , Subtraction , Multiplication , and Division .
What improvements may be had from hence in the Extraction of Roots , whether of pure or affected Equations , I shall leave to future inquiry .
But all these Operations , whether in the common Decimal Arithmetic , or in this here compendiously described ; and not only in Decimal Arithmetic , but in the several other Species , as Duodecimal , Sexage simal , Centesimal , &amp;c. all these Operations ( I say ) may be much more easily and readily performed , and with equal accuracy , by an Instrument I have contrived called Abacus , or the Counting-Table , which I hope shortly to communicate to the inquisitive in these matters , and by which all long Calculations may be very much facilitated .
In the mean time this short account of Negativo-affirmative Arithmetic , in Denary or Decimal Numbers , may be premis 'd , by way of Introduction to the knowledge and use of the said Instrument

