Philos. Trans. Vol. LV .
TAB. VI.p .
205 .
P E II D S Fig. 2 .
L Fig. 3 .
CH FA C A G BK F DB S L Fig. 4 .
AaC EB Se L l Fig. 5 .
AC B Philos. Trans. Vol. LV .
TAB. V. p. 205 .
Fig. 1 .
70 60 50 40 30 20 10 E C RR KLCL N B P H L K G F From a Scale of 3,63 Inches to a Foot .
XXV .
An Account of a Balance of a new Construction , supposed to be of Use in the Woollen Manufacture .
By W. Ludlam , B. D. Fellow of St. John 's College , Cambridge .
IT is of consequence in some branches of the wollen manufacture , that the thread of which any piece is woven should be all of the same fineness .
After it is spun , it is made into skains of the same length , and these are sorted according to the fineness of the spinning .
The manufacturers usually distinguish and denominate the fineness , by the number of skains which go to the pound ; the coarsest being about 12 to the pound , and the finest near 60 .
There is no other method of sorting in use , except by the eye ; but it requires great nicety to distinguish the size of threads so small , and long experience to know by the look only , how many skains of any particular sort will make a pound .
A method of weighing them readily would save much time : the machine here delineated is for that purpose .
It resembles the beam of a common pair of scales ; at one end of it is a fixed weight , which I call the counterpoise , at the other a hook : in sorting , the skain to be examined is put upon the hook , and sinks down more Read June 6 , 1765 .
FIG. 1 .
more or less , according to its weight , till the counterpoise by rising balances it , and then the index or cock of the beam points out , on a graduated arch , the number of skains of that sort , which go to the pound .
A scale , instead of the hook , might be used for weighing money , if the arch were properly divided for that purpose .
Mr. Rouse of Harborough , many years ago , made a machine for sorting woollen thread upon the same principle with this ; but as what he did was mostly tentative , he was not aware of some considerable advantages which the theory points out .
For the machine will not distinguish with equal nicety the skains of every size .
In Mr. Rouse 's machine , the divisions were too small , and the largest chanced to fall at 18 to the pound ; but it would have been better if the finer sorts had been more accurately distinguished , as being of greater consequence to be well sorted , and more difficult to be sorted by the eye only , than the coarser ones .
This machine distinguishes best the yarn of 36 to the pound , one of the finest sorts , as I am informed , in common use , the largest division lying between 36 and 37 ; the other divisions are as large , and the whole range of the index as much as can be allowed without other inconveniences .
The theory contains the necessary rules for finding the angle of the beam , for calculating the divisions on the arch , and for placing their largest interval in any part of them .
Directions for making the Balance .
IT consists of a mahogony stand , a steel beam and brass ring for the divisions .
FGH is the triangular base of the stand , having a screw in each angle to set it level : into this is dove-tail 'd the upright back K K ; the standard board LLL is put into a dove-tail 'd groove in the back , and tenonted into the triangular board at the bottom .
FIG. 1 .
The two cocks C C , between which the arbor of the beam plays , and the ring RR are screwed to the standard board L L. The beam A B , with its cock or index E , is broad in the middle to gain strength , and pierced to make it lighter .
It is rivetted on a collet soldered to the arbor , as clock wheels are .
The pivots run in hard brass , and have plates of hardened steel for their points to bear against .
At one end of the beam is the counterpoise A , consisting of two round pieces of brass , screwed together through an hole in the beam .
The other end of the beam is made thicker for about half an inch in length , and slit to receive a loop of hardened steel , which hangs on the steel pin B ; on the lower part of the loop hangs the hook , which holds the skain .
The block of wood N , is screwed to the standard board ; the upper part of which is lined with velvet for the counterpoise to rest upon , when the skain is taken off the hook ; the brass pin P is for the other arm of the beam to bank against .
The angle of the beam ACB is 168~ 6~ , the radius CA or C B , 6 inches ; the breadth of the beam in the middle 3/ 4 of an inch , at the ends 1/ 4 ; the thickness at the end B where it is slit 1/ 6 ; every where else 1/ 18 .
The length of the arbor 1,3/ 4 ; diameter of the pivets 1/ 30 .
The weight of the counterpoise one ounce avoirdupois .
The hook , with the loop and steel pin included , 0,68 avoirdupois .
The division of 50 is at the top of the ring .
In making the beam , the point of the index E must be equally distant from the centers A and B ; the whole beam and index made so as to poise itself , and remain at rest in any position , before the counterpoise steel pin and loop are put on .
The counterpoise being at first made too heavy , to adjust it , take off the beam , but let the two cocks and ring remain screwed on ; hang then a fine wire and plummet on the top division , and with a watch glass look through one pivot hole till you see the wire against the other ; turn the screw G till the wire bisects the hole , and the top division is then perpendicularly over it .
Take away the plumb line and put on the beam and counterpoise ; on the loop hang a weight , which together with the steel pin and loop makes one ounce avoirdupois ; turn now , by degrees , some metal off the counterpoise , till the index points at the top division , and the counterpoise will be truly adjusted , whether the arms of the beam , C A , C B , are precisely equal or not : for though the arms should be nearly equal , it is not necessary they should be exactly so , as in scales .
The divisions are set on the ring by an instrument made on purpose , which will very readily cut them , tho ' unequal , with great exactness , on any circle large or small .
To prove the beam , put a weight into the scale ( that of 28 to the pound is the best in this case ) and see if it brings the index down to the proper division on the ring : if it carries it too far , the angle of the beam ACB is too great ; if the contrary , too little ; and the arms C A , C B , must be set a little in , or out , till the angle is right : or the angle of the beam may be first found experimentally , by the rule hereafter given , and the divisions calculated to it , which is not much trouble ; for having a table ready made for the intended angle , the alterations in that table occasioned by a small variation from that angle will be easily found by the rules at the end of the following Theory .
When the balance is to be used , a weight of 1/ 50 of a pound avoirdupois is to be put on the hook : The screw G must then be turned , till the index E points at the division of 50 : The machine is then properly adjusted , and the weight may be taken off , and a skain put on .
Let ACB be a bent lever moveable about the angular point C , Ba scale at the end of the arm B , Aa counterpoise at the end of the other arm A ; Given the angle of the lever , the length of the arms , the respective weights of the scale and counterpoise , to find the position of the lever when at rest .
FIG. 2 .
Produce the arm BC to I , so that CI may be to CB as the weight of the scale to the counterpoise ; join A I , turn the lever about till AI is perpendicular to the horizon , and in that position it will be at rest .
For if HB be drawn parallel to I A , and FCH perpendicular to HB and I A , then C F : C H : C I : C B , or as the scale to the counterpoise ; that is , perpendiculars from the center of motion to the lines of direction are reciprocally as the forces applied , therefore in that case the forces balance each other .
Cor. Join AB and draw CG parallel to A I , and A G : G B : : C I : C B , or as the scale to the counterpoise , therefore G is their center of gravity , which lying in a perpendicular to the horizon passing thro ' the center of motion , again shows the lever to be at rest .
SCHOLIU M. There are indeed two positions in which the lever will be at rest , one when the center of gravity is above , and another when it is below the center of motion ; the latter only is considered in what follows .
Let the arms CA and CB be equal , let the weight of the counterpoise be fixed , but that of the scale variable ; then if the fixed line CA ( = C B ) represent the former , the variable line CI will represent the latter .
In the line C I , take CD =CA and join A D ; thro ' the center C , draw ECK parallel to A D , which therefore bisects the angle of the lever AC B , and the part EC may stand for the index or cock of the lever .
Let CP be perpendicular to the horizon , and ECP the angle of the index with that perpendicular is equal to DA I ; but ACK is the semi-sum , and DAI the semi-difference of the angles angles CA I , CI A : Therefore CI+C A : CIC A : : or the sum of the weight of the scale and counterpoise is to their difference , as the tangent of AC K , half the angle of the lever , to the tangent of DAI =EC P , the inclination of the index to the perpendicular when the lever is at rest : and the angles at A and I , or the inclinations of the arms , will be the sum and difference of ACK and DA I. Hence the scale and counterpoise being given , the inclination of the index may be found from the angle of the lever ; or the angle of the lever from the inclination of the index .
The weight of the counterpoise being fixed , let that of the scale vary uniformly ; it is required to find when the angular motion of the index is greatest .
All things remaining as before , draw AS perpendicular to C I ; and the variation of CI will be the same as that of SI ( the line CS being constant ) : For a like reason the variation of the angle DAI is the same with that of SA I. Now the variation of S I , is to that of the arch which measures SA I , ( A S being radius ) as AI2 to AS 2 ; if therefore CI or SI flows uniformly , the fluxion of SAI or DA I , or the angular motion of the index , will be greatest when AI is least , that is , when AI coincides with A S , or when the inclination of the index is equal to DA S , or when the arm carrying the scale is parallel to the horizon .
The angle DAS is the complement of ADS ( = AC K ) or of half the angle of the lever .
Again C A : C S : or radius , to the co-sine of the angle of the lever as the counterpoise to the weight of the scale , when the angular motion of the index is greatest .
Let now the weight of the scale as well as the counterpoise be fixed , but into it let there be successively put variable weights or skains of thread , to be denominated by the number of them which ( at that size ) go to the pound : Let the weight of these skains vary at each successive change in such a manner , that the number which goes to the pound may be increased each time by unity : It is required to find when the angular variation of the index , in one exchange , will be greatest .
All things as before , let CL represent the fixed weight of the scale , LI the variable one of the skain , CI that of both together .
Call A S , s : S L , d : put p= one pound , x the number by which any one skain is denominated , consequently p/ x its weight =L I ; whence SI=L I-L S= p/ x-d , whose fluxion ( putting the uniform fluxion of x to be unity ) is -p/ xx : but the fluxion of SI is to that of the arch which measures SA I , as AI2 to AS 2 , or as AS2+ SI2 to AS 2 , or as ss + p/ x-d 2 : ss , that is ss+p/ x-d 2 : ss : : -p/ xx to the variation of the index sought ; whence that variation is greatest when x= pd/ ss+dd , that is when p/ x or the weight of the skain = ss+dd/ d : wherefore if CA represent the counterpoise , and CL the scale , join AL and draw AF perpendicular FIG. 3 .
pendicular to it , cutting CL produced in F , then will LF represent the weight of the skain , and DAF be the inclination of the index , when its variation is greatest .
On FL as a diameter describe a semicircle , and let CA intersect it again in a : Join L a , a F , then will LaF be a right angle as well as LA F ; therefore with the given angle of the lever ACB and given scale C L , we have the same maximum F L , either with the counterpoise C A , or C a ; either of which gives the other ; for C A : C L : : C F : C a. FIG. 4 .
In like manner , if the angle of the lever , the counterpoise C A , and FL the weight of the skain at the maximum are given , with the center A and radius AE= 1/ 2 FL describe a circle cutting BC produced in E and e , towards C set off EL and e l , equal to the radius aforesaid , and either CL or Cl will be the weight of the scale that will give the proposed maximum .
It is manifest that if AE or 1/ 2 FL is less than A S , the problem is impossible ; therefore the maximum is limited , and must be proposed such , that the weight of the skain in that case be less than 2A S. FIG. 4 .
FIG. 5 .
EXAMPL E : Let the semi-angle of the lever be 84~ : 03~ , the counterpoise one ounce , let the division for 50 to the pound , stand at the top of the graduated arch perpendicularly over the center of the lever , that is , let LD= 16/ 50 oz. consequently LC or the scale =CDL D , or CALD=1 16/ 50 = 2 0,68 0,68 oz. then is CS= 0,9785090 , LS or d= 0,2985090 , AS or s= 0,2062042 , and ss+dd/ d or ss/ d +d ( = AS 2/ L S+L S ) =LF= 0,440951 oz. and 16/ 0.440951 = 36,2852 ; therefore the greatest motion of the index will be when a skain of 36 to the pound is changed for one of 37 , and the largest interval on the graduated arch will be that which lies between 36 and 37 .
C A : C L : : C F : Ca or 1 : 0,68 : : 1,120951 : 0,762247 ounces , the other counterpoise which gives the same maximum .
Let the maximum be assumed at 36 skains to the pound exactly , or when FL= 0,4444444 , then is 1/ 2 FL or AE= 0.2222222 , and SE ( = ~AE+ASXAEA S ) = 0,0828404 , whence CL ( = CSAE+S E ) = 0,8391272 , and CL ( = CSAES E ) = 0,6734464 : if this last quantity be taken for the weight of the scale , the division for 49 in the pound would fall at the top of the arch very nearly , therefore assume the division of 50 in the pound for the top , which will make but little alteration in the maximum , and the weight of the scale will be 0,68 , the maximum between 36 and 37 , as before .
2AS= 0,4124084 , and 16/ 0.4124084 = 48 , 8419 ; therefore with the angle of the lever and counterpoise as before , the maximum cannot be higher than at 48 skains to the pound .
In Mr. Rouse 's machine the the angle of the lever was 165~ , the counterpoise 1.46057 ounces , the scale 1,19543 or AS= 0,378024 , therefore his machine did not admit of a maximum higher than between 26 and 27 skains to the pound .
To give one example of finding the inclination of the index , let it be required to find the angle for a skain of 32 to the pound , whose weight is 0 .
5 ounce ; whence CI= 1.18 , CA= 1 , and CI+C A : CIC A ; or 2,18 : 0,18 : : Tang , ACK= 84~ : 03~ .
Tang. ECP= 38~ : 23~ .
FIG. 3 .
This rule for the inclination of the index , suggests an easy way of finding the alterations in it , occasioned by any small variation in the angle of the lever , other things remaining the same .
For the change in the Logarithmic Tangent of EC P , the fourth term in the proportion , is equal to that made in the Logarithmic Tangent of AC K , the third term ; the other two terms being the same .
Take therefore the difference of the Logarithmic Tangents of the two semi-angles , and divide it by the natural numbers 1 , 2 , 3 , 4 , &amp;c. successively , and their quotients among the differences of the Logarithmic Tangents , in Tables for each minute , will point out the quantity of the angle EC P , when the alteration in it is 1 , 2 , 3 , 4 , &amp;c. minutes successively ; by which means a table may be easily formed for the correction of that first computed .
Suppose , for instance , the semi-angle of the lever to be only 84~ : instead of 84~ : 03~ , for which a table has been made , it is required to correct this table , so as to suit it to this semi-angle 84~ .
The difference of the Log. Tangents of these semi-angles is 36.608 ; divide this by any of the natural numbers , for instance by 10 , the quotient 3660 , standing among the Tabular differences ( in Tables for each minute ) at 21~ : 50~ and 68~ : 10~ , shows that when the value of ECP in the table already made , is either 21~ : 50 or 68~ : 10~ , 10 minutes must be substracted to adapt it to the lever whose semi-angle is only 84~ : and so of the rest .
The greatest difference of the inclination of the index will be when ECP is , 45~ : the Tabular differences being then the least .
All this is not indeed quite accurate , but sufficiently so for practice .
To find accurately when this difference is greatest , and what that greatest difference is ; add ( or subtract as the case requires ) half the difference of the Logarithmic Tangents of the semi-angles aforesaid , to the Logarithmic Radius , and it will give the Logarithmic Tangent of EC P , in that case ; and the difference between ECP thus found , and its complement , is the greatest difference sought .
Thus half 36,608 added to 10.0000000 is 10.0018304 , the Log. Tangent of 45~ : 7~ : 14~ , 4 , the inclination of the index , when its change by reducing the semi-angle of the lever to 84~ , is greatest , and the quantity of that change is 14~ : 28~ , 8 .
A Table of the inclination of the index , when the semi-angle of the lever is 84~ : 03~ , the counterpoise ~ ounce and the scale 0 .
68 ounces avoirdupois , for any number of skains to the pound from 10 to 70 .
Skains to the pound .
~'~'~~ ' 10 75 : 03 25 52 : 56 40 20 : 15 55 8 : 04 11 73 : 56 26 51 : 00 41 18 : 02 56 9 : 30 12 72 : 47 27 49 : 01 42 15 : 50 57 10 : 53 13 71 : 35 28 46 : 59 43 13 : 41 58 12 : 13 14 70 : 20 29 44 : 54 44 11 : 35 59 13 : 30 15 69 : 02 30 42 : 46 45 9 : 31 60 14 : 44 16 67 : 40 31 40 : 36 46 7 : 30 61 15 : 55 17 66 : 16 32 38 : 23 47 5 : 32 62 17 : 03 18 64 : 48 33 36 : 09 48 3 : 38 63 18 : 08 19 63 : 17 34 33 : 54 49 1 : 47 64 19 : 11 20 61 : 42 35 31 : 37 50 0 : 00 65 20 : 12 21 60 : 04 36 29 : 20 51 1 : 44 66 21 : 10 22 58 : 22 37 27 : 03 52 3 : 24 67 22 : 06 23 56 : 37 38 24 : 46 53 5 : 01 68 23 : 00 24 54 : 48 39 22 : 30 54 6 : 34 69 23 : 52 70 24 : 41 N. B. According to the report from the committee appointed by the House of Commons to enquire into the original standard weights , one pound Avoirdupois is 1 lb. 2 oz. 12 dwt .
Troy ; consequently one ounce is 18 dwt .
6 gr. and 0.68 oz. is 12 dwt .
9.84 grains .

