LXIX .
A Letter from Mr. J. Smeaton to Mr. John Ellicott , F. R. S. concerning some Improvements made by himself in the AirPump .
SIR , I HAVE been informed by some of my friends , that my endeavours towards completing the air-pump , have been mentioned with approbation , in papers that Mr. Short and Mr. Watson have lately communicated to the Royal Society .
I understand likewise , that the latter of those gentlemen has , in a very obliging manner , expressed an inclination , that I should lay before them a particular account of my improvements therein .
Read April 16 , 1752 .
I shall always esteem it a singular honour to be thought capable of producing any thing worthy the attention of the Royal Society ; and to be my duty and interest so to do , upon the least intimation of that kind .
Your superior skill in mechanics , together with the assistance you have given me in making trial of my pump , against three very good ones of the common construction , as well as the frequent marks of friendship you have shown me on all occasions , encourage me to trouble you with communicating the following to that Society , of which you are a member , and who , of all others , are the most proper judges .
I shall not take up time with a particular recital of the alterations I have made , for near four years past , in order to remove some obstacles , which I imagined hindred the effects , that the theory I set out upon seemed to promise .
It will be sufficient , that I give an account of what has appeared to answer best , after a great number of different trials ; which , tho ' short of what I at first expected , yet as this pump performs much better than the common ones , my labour may not be thought wholly useless ; and the respect , which I have to the Society , would still have prevented me from troubling you or them about it at this time , could I have thought of any alteration , that promised materially to improve it .
The principal causes of imperfection in the common pumps arise , first , from the difficulty in opening the valves at the bottom of the barrels ; and , 2dly , from the piston 's not fitting exactly , when put close down to the bottom ; which leaves a lodgement for air , that is not got out of the barrel , and proves of bad effect , as I shall show in the course of this paper .
In regard to the first of these causes ; the valves of air-pumps are commonly made of a bit of thin bladder , stretched over a hole generally much less than one tenth of an inch diameter ; and to prevent the air from repassing between the bladder and the plate , upon which it is spread , the valve must always be kept moist with oil or water .
It is well known , that at each stroke of the pump the air is more and more rarefied , in a certain progression , which would be such , that an equal proportion of the remainder would be taken away , was it not affected by the impediments I have mentioned : so that , when the spring of the air in the receiver becomes so weak , as not to be able to overcome the cohesion of the bladder to the plate , occasioned by the fluid between them , the weight of the bladder , and the resistance that it makes by being stretched , the rarefaction cannot be carried farther , tho ' the pump should still continue to be worked .
It is evident , that the larger the * hole is , over which the bladder is laid , a proportionably greater force is exerted upon it by the included air , in order to lift it up ; but the aperture of the hole cannot be made very large , because the pressure of the incumbent air would either burst the valve , or so far force it down into the cavity , as to prevent its lying flat and close upon the plate , which is absolutely necessary .
To avoid these inconveniences as much as possible , instead of one hole , I have made use of seven , all of equal size and shape ; one being in the centre , and the other six round it : so that the valve is supported at proper distances , by a kind of grating , made by the solid parts between these holes : And to render the points of contact , between the bladder and grating , as few as possible , the holes are made hexagonal , and the partitions filed almost to an edge .
As the whole pressure of the atmosphere can never be exerted upon this * If we examine the force , that air rarefied 140 times can exert in a common valve through a hole of one tenth of an inch diameter , we shall find it not to exceed 6 grains at a medium .
this valve , in the construction made use of in this pump ; and as the bladder is fastened in four places instead of two , I have made the breadth of the hexagons three tenths of an inch ; so that the surface of each of them is more than nine times greater than common .
But as the circumference of each hole is more than three times greater than common , and as the force , that holds down the valve , arising from cohesion , is , in the first moment of the air 's exerting its force , proportionable to the circumference of the hole ; the valve over any of these holes will be raised with three times more ease than common .
But as the raising of the valve over the center-hole is assisted on all sides by those placed round it ; and as they all together contribute as much to raise the bladder over the center-hole , as the air immediately acting under it ; upon this account the valve will be raised with double the ease , that we have before supposed , or with a sixth part of the force commonly necessary .
It is not material to consider the force of the cohesion , after the first instant : For , after the bladder begins to rise , it exposes a greater surface to the air underneath , which makes it move more easily .
I have not brought into this account the force , that keeps down the valve , that arises from the weight of the bladder , and the resistance from its being stretched ; for I look upon these as small , in comparison of the other .
I was not however contented with this construction of the valves , till I had tried what effect would be produced , when they were opened by the motion of the winch , independent of the spring of the air : And tho ' the contrivance I made use of seemed to me less liable to objection than any thing I was acquainted with , that had been designed for that purpose ; yet I did not find it to answer the end better than what I have already described ; and therefore laid it aside , as it rendered the machinery much more complex , and troublesome to execute .
But supposing all those difficulties to be absolutely overcome , the other defect , that I mentioned in the common construction , would hinder the rarefaction from being carried on beyond a certain degree .
For , as the piston cannot be made to fit so close to the bottom of the barrel , as totally to exclude all the air ; as the piston rises , this air will expand itself ; but still pressing upon the valve , according to its density , hinders the air within the receiver from coming out : Hence , were this vacancy to equal the 150th part of the capacity of the whole barrel , no air could ever pass out of the receiver , when expanded 150 times , tho ' the piston was constantly drawn to the top ; because the air in the receiver would be in aequilibrio with that in the barrel , when in its most expanded state .
This I have endeavoured to overcome , by shutting up the top of the barrel with a plate , having in the middle a collar of leathers , through which the cylindrical rod works , that carries the piston .
By this means , the external air is prevented from pressing upon the piston ; but that the air , that passes thro ' the valve of the piston from below , may be discharged out of the barrel , there is also a valve applied to the plate at the top , that opens upwards .
The consequence of this construction is , that when the piston is put down to the bottom of the cylinder , the air in the lodgment under the piston will evacuate itself so much the more , as the valve of the piston opens more easily , when pressed by the rarefied air above it , than when pressed by the whole weight of the atmosphere .
Hence , as the piston may be made to fit as nearly to the top of the cylinder , as it can to the bottom , the air may be rarefied as much above the piston , as it could before have been in the receiver .
It follows therefore , that the air may now be rarefied in the receiver , in duplicate proportion of what it could be upon the common principle ; every thing else being supposed perfect .
Another advantage of this construction is , that tho ' the pump is composed of a single barrel* , yet the pressure of the outward air being taken off by the upper plate , the piston is worked with more ease ~ than the common pumps with two barrels : And not only so , but when a considerable degree of rarefaction is desired , it will do it quicker ; for the terms of the series expressing the quantity of air taken away at each stroke do not diminish so fast , as the series answering to the common one .
I have found the gages , that have been hitherto made use of , for measuring the expansion of the air , very unfit to determine in an experiment of so much nicety .
* It is obvious that these improvements will equally obtain , whether the pump is constructed with a single or a double barrel .
~ Because , tho ' the pressure of a column of air , equal to the diameter of the piston-rod , still presses upon it , yet , as there is only the friction of one piston , and that not loaded with the weight of the atmosphere ; the friction of the leather against the side of the barrel , and that of the rack and wheel , is much less : so that , notwithstanding the addition of friction in the collar of leathers , that of the whole will be less .
nicety .
I have therefore contrived one of a different sort , which measures the expansion with certainty , to much less than the 1000th part of the whole .
It consists of a bulb of glass something in the shape of a pear , and sufficient to hold about half a pound of quicksilver .
It is open at one end , and at the other is a tube hermetically closed at top .
By the help of a nice pair of scales , I found what proportion of weight a column of mercury , of a certain length , contained in the tube , bore to that , which filled the whole vessel .
By these means I was enabled to mark divisions upon the tube , answering to a 1000th part of the whole capacity , which being of about one tenth of an inch each , may , by estimation , be easily subdivided into smaller parts .
This gage , during the exhausting of the receiver , is suspended therein by a slip-wire .
When the pump is worked as much as shall be thought necessary , the gage is pushed down , till the open end is immerged in a cistern of quicksilver placed underneath : The air being then let in , the quicksilver will be driven into the gage * ; till the air remaining in it becomes of the same density with the external ; and as the air always takes the highest place , the tube being uppermost , the expansion will be determined by the number of divisions occupied by the air at the top .
The degree , to which I have been able to rarefy the air in experiment , has generally been about 1000 times , * The bulb of the gage may be emptied of its quicksilver , without taking that out of the tube ; and the tube being held horizontal , the column of mercury in it will have power to contract or expand the air at the top .
times , when the pump is put clean together : But the moisture , that adheres to the inside of the barrel , as well as other internal parts , upon letting in the air , is in the same succeeding trials worked together with the oil , which soon renders it so clammy , as to obstruct the action of the pump upon a fluid so subtle as the air is , when so much expanded ; but in this case it seldom fails to act upon the air in the receiver , till it is expanded 500 times : And this I have found it to do , after being frequently used for several months , without cleaning .
I have also generally found it to perform best , the first trial at each time of using ; tho ' nothing had been at it from the time preceding ; which , after a great many trials made with this view , I also attribute to the vapours of the air mixing with the oil .
An experiment , where the air was expanded 1000 times , was tried about two years since in your presence ; at which were present also Dr. Knight and Mr. Canton ; and I lately did the same thing with Mr. Watson .
The pump , which I intend myself the honour of showing the Society , is the same , that I just now mentioned , and the second that I made , with a view to improve upon this principle .
The degree of rarefaction , produced by the best of the three pumps , that you procured the trial of , and which you esteemed good in their kind , and in complete order , never exceeded 140 times , when tried by the gage above described .
I have also endeavoured to render the pneumatic apparatus more simple and commodious , by making this air-pump act as a condensing engine at pleasure , by singly turning a cock .
This not only enables us to try any experiments under different circumstances of pressure , without changing the apparatus , but renders the pump an universal engine , for showing any effect , that arises from an alteration in the density or spring of the air .
Thus , with a little addition of apparatus , it shows the experiments of the air-fountain , wind-gun , &amp;c. This is done in the following manner : The air above the piston being forcibly driven out of the barrel at each stroke , and having no-where to escape , but by the valve at the top ; if this valve be connected with the receiver , by means of a pipe , and at the same time the valve at the bottom , instead of communicating with the receiver , be made to communicate with the external air , the pump will then perform as a condenser .
The mechanism is thus ordered .
There is a cock with three pipes placed round it , at equal distances .
The key is so pierced , that any two may be made to communicate , while the other is left open to the external air .
One of these pipes goes to the valve at the bottom of the barrel ; another goes to the valve at the top , and a third goes to the receiver .
Thus , when the pipe from the receiver , and that from the bottom of the barrel , are united , the pump exhausts : But turn the cock round , till the pipe from the receiver , and that from the top of the barrel , communicate , and it then condenses .
The third pipe , in one case , discharges the air , taken from the receiver , into the barrel ; and in the other , lets it into the barrel , that it may be forced into the receiver .
I am , SIR , Furnival's-Inn-court , Your most humble servant , April 16 , 1752 .
J. Smeaton .
P. S. I have also added some draughts , and letters of reference , in order to explain myself more fully .
Figure I. Is a perspective view of the principal parts of the pump together .
A is the barrel .
B the cistern , in which are included the cock , with several joints .
These are covered with water to keep them air-tight .
A little cock to let the water out of the cistern , is marked 6 .
Ccc is the triangular handle of the key of the cock : which , by the marks on its arms , shows how it it must be turned , that the pump may produce the effect desired .
DH is the pipe of communication between the cock and the receiver .
E is the pipe , that communicates between the cock and the valve , on the upper plate of the barrel .
F is the upper plate of the pump , which contains the collar of leathers d , and V the valve , which is covered by the piece f. K Fig. 4 .
p. 427 A E DB I F Fig. 2 .
p. 425 .
C KO LMP HNQP IA Fig. 1 .
b p. 424 .
a G I F EC D AA C~VV ~~X P. bW GGBS H T TH S Philos. Trans. Vol. XLVII .
TABXVIII .
p. 427 .
A Fig. 6 .
H p. 494 .
K D D C GI is the siphon-gage ; which screws on and off , and is adapted to common purposes .
It consists of a glass tube hermetically sealed at c , and furnished with quicksilver in each leg ; which , before the pump begins to work , lies level in the line a b ; the space bc being filled with air of the common density .
When the pump exhausts , the air in bc expands , and the quicksilver in the opposite leg rises , till it becomes a counter-balance to it .
Its rise is shown upon the scale I e , by which the expansion of the air in the receiver may be nearly judged of .
When the pump condenses , the quicksilver rises in the other leg , and the degree may be nearly judged of by the contraction of the air in b c : marks being placed at 1/ 2 and 1/ 3 of the length of bc from c ; which shows when the receiver contains double or treble its common quantity .
KL is a screw-frame to hold down the receiver , in condensing experiments , which takes off at pleasure ; and is sufficient to hold down a receiver , the diameter of whose base is 7 inches , when charged with a treble atmosphere : in which case it acts with a force of about 1200 pounds against the screw-frame .
M is a screw , that fastens a bolt , which slides up and down in that leg , by means whereof the machine is made to stand fast on uneven ground .
Fig. II .
Is a perpendicular section of the barrel and cock , &amp;c. where AB represents the barrel .
CD the rod of the piston , which passes through MN the plate , which closes the top of the barrel .
K is the collar of leathers , through which the pistonrod passes .
When the piston is at the bottom of the cylinder , the upper part of K is covered by the cap at D , to keep out dust , &amp;c. L is the valve on the upper plate , which is covered by the piece O P , which is connected with the pipe Q R , which makes the communication between the valve and cock .
CE is the piston ; and EFF is the piston-valve .
II are two little holes to let the air pass from the piston-valve into the upper part of the barrel .
GGK is the principal valve at the bottom of the cylinder .
HH is a piece of metal , into which the valve GGK is screwed , and closes the bottom of the cylinder ; out of which also is composed SS the cock , and KTT the duct from the cock to the bottom of the barrel .
WW is the key of the cock .
X the stem ; and W the handle .
Fig. III .
Is an horizontal section of the cock , through the middle of the duct TT .
AB represents the bigness of the circular plate , that closes the bottom of the barrel .
CD represents the bigness of the inside of the barrel .
EFG is the body of the cock ; the outward shell being pierced with 3 holes at equal distances , and corresponding to the three ducts HH , II , KK , whereof HH is the duct , that goes to the bottom of the barrel .
I I , the duct , that communicates with the top of the barrel ; and KK is the duct , that passes from the cock to the receiver .
LMN is the key , or solid part of the cock , moveable round in the shell EFG .
When the canal LM answers to the ducts HH and KK , the pump exhausts , and the air is discharged by the perforation N. But the key LMN being turned till the canal LM answers to II and KK , the perforation N will then answer to HH ; and in this case the pump condenses .
Lastly , when N answers to KK , the air is then let in or discharged from the receiver , as the circumstance requires .
Fig. IV .
Is the plan of the principal valve .
ABCD represents the bladder fastened in 4 places , and stretched over the 7 holes IK , formed into an hexagonal grating ; which I shall call the honeycomb .
EFGH shows where the metal is a little protuberant , to hinder the piston from striking against the bladder .
Fig. V. Represents the new gage ; which I call the peargage .
It is open at A ; BC is the graduated tube , which is hermetically closed at C , and is suspended by the piece of brass DE , that is hollowed into a cylinder , and clasps the tube .

