3 58
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
mens of several bronzes, Monel metal, and steel were weighed and embedded in rich earth, which was kept wet for six months b y periodical additions of very dilute solutions of corrosive salts. At the end of the test period all the specimens were taken out, scrubbed, dried and weighed t o ascertain the comparative loss from corrosion. The results were as follows: Phosphor bronze.. .................... Tobin bronze..
Per cent. loss. 0.09 0.11 0.12 0.12 0.33 1.04
....................... ........... ........................
........................
Monel metal., Parsons manganese bronze.. Muntz metal., Steel ................................
Another test of the same kind, under somewhat different conditions, but the same period, gave about the same relative results. One advantage in favor of Monel metal in these tests was that it presented the least change in ap.pearance as a result of the corrosive action. CASTING.
'
While no trouble has been experienced in rolling Monel metal, considerable difficulty has been met with in some foundries rn getting sound castings. Incidentally, i t may be remarked that nickel-copper alloys have a great tendency to develop blow-holes on casting, due t o occluded gases. This is another indication of the absence of a chemical compound, since the presence of a compound in a n alloy seems t o have the influence of preventing blow-holes, and closing the grain. Monel metal requires special precautions in casting on account of dissolved oxides and gases, which, if not removed, would render the ingots unsound. Successful castings have been made in small foundries by adding z ounces of magnesium per IOO Ibs. of the alloy, before pouring. Unsound castings may a?so arise from gases that are liberated in the cores of the patterns, and from the fact that the freezing point of the metal is so many hundred degrees above that of brass or bronze t h a t these gases cannot escape before they are trapped b y the freezing metal. This necessitates the use of the same methods as for fine steel castings. I t must be remembered that the actual temperature of Monel metal cannot be much lower than 1550' C. The art of casting Monel metal is closely related to the a r t of casting steel, as distinct from the art of casting brass, bronze or cast iron. It really involves the art of handling the finest steel castings. The practice of casting Monel metal is, therefore, very similar t o t h a t followed in steel foundries. The melting equipment of the plant consists of standard reverberatory furnaces which are fired with oil, the fuel being fed b y gravity. The metal is charged into the furnace in the form of pigs or slabs which are about 15 X 8 X 3 inches in dimensions, and weigh about 85 lbs. The melting point is about 2j00' F. and eight hours are required t o the heat. No fluxes nor alloys are necessary. Because of the relatively scant knowledge of nickel and its peculiarities, the handling of Monel metal requires expert attention in the melting. The industry of making Monel metal castings is
May, 1912
now past the experimental stage, and is being SUCcessfully accomplished in a t least three foundries in New York and New Jersey, Mr. Leonard Waldo, consulting engineer, is authority for the statement that notable success has been achieved a t the Johnson Foundry of Spuyten Duyvil in the casting of this metal. The Bayonne Casting Company, of Bayonne, N. J., and the Riverside Steel Company, of Newark, N. J., are other foundries that are making good castings. USES.
Although a comparatively new alloy, because of its resistance to corrosion, its strength and the readiness with which i t may be machined, Monel metal has already found many important uses. These include pump cylinders for handling salt water, propeller wheels, rudders, centre boards, deck fittings, mining screens, water ends of pumps, linings, valves, shafts, piston rods and plumbing fixtures subject t o corrosive influences. I n the rolled metal, there is a wide range of usage. Striking instances in the past include the roofs of the new terminals of the Pennsylvania Railroad in New York, and the Chicago & Northwestern Railroad in Chicago, which were covered with Monel metal. The sheets have also been used on s'everal large office buildings in New York, and on industrial plants subject t o severe corrosive conditions where copper had proved unsatisfactory. Rigid physical specifications should be the rule wherever Monel metal is considered for use in any form of engineering construction. With reference t o chemical specifications, i t would be impracticable t o set any narrow limits as to chemical constituents, owing to the somewhat variable character of the ore and the method of manufacture. It is not believed, moreover, that such limits are necessary in order t o insure the desired physical properties of strength and durability. If too close limits are set on chemical constituents, the metal so specified might be obtained, but a t a cost that would render its use prohibitory. LABORATORY, BOARD OF WATERSUPPLY, NEWYORKCITY.
FREE LIME IN PORTLAND CEMENT.' By H. E. KIEFER.
Received January 2, 1912.
I t is to be regretted that the majority of chemists engaged in the manufacture of Portland cement have very little time for research work. Some of us have considerable time for experimental work, but the time and money spent are expected t o produce commercial results rather than contribute t o the stock of purely scientific knowledge. This latter we are compelled to leave t o the research chemist, and as he usually lacks a practical manufacturing experience we may be able t o supply a few facts and suggestions. One of the most troublesome arguments which the cement chemist has t o meet is on the subject of Free Lime. We hear i t discussed b y chemists and engi1
Read at the Forty-fifth annual meeting of the A. C. S., at Washing-
ton, December, 1911.
May, 1912
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
neers and attribute all kinds of defects in the cement to it, yet it is very questionable whether any of us have a well grounded theory for the position we assume. We manufacture a product to meet certain specifications and when it fails in any of these we look for the cause and a remedy from a practical rather than a theoretical point of view, using theory only to facilitate results. A few things we know and many more we believe, hence we shall state first the general knowledge and follow it up with the theory a n d then add a few observations which may furnish suggestions for research chemists. Constancy of volume is one of the requirements of practically all specifications. The test is made b y subjecting a hardened p a t t o five hours’ boiling in water in a loosely covered vessel or t o the steam above the water in such a vessel. If the pat comes out sound and hard, with no signs of warping, checking or disintegration, the cement is considered sound. On the other hand, if it becomes soft, or warps or cracks on the edges it is considered unsound. From this point the theory begins. The explanation is t h a t a failure to pass the test is due t o free lime. Various quantitative methods have been suggested with a view of determining it, but so far as the writer knows none has ever given satisfaction. Microscopic tests have been suggested, but these likewise fail to be of any practical value. Some observations made in our laboratory lead t o the belief that if free lime is a real cause i t is not the only one, as there is a physical side which must be taken into consideration. Le Chatelier says that if one per cent. of freshly calcined lime be added t o a sound cement it will render it unsound. With this as a starting point, five well known American brands of Portland cement, which passed the boiling test as received, were taken and the above percentage of lime, freshly calcined over a blast lamp, immediately pulverized and added to the cement. None of them failed on the test. Larger quantities were than added, and with one brand 2 5 % of freshly calcined free lime was required before the cement failed on the boiling test. The authority mentioned fails to say how finely the lime should be ground before adding to the cement, but i t was presumed i t should be a t least as fine as the cement, hence i t was quickly put through a Ioo-mesh sieve. The lime was obtained from a pure limestone a t the highest temperature of a blast lamp. It might have been better t o have used calcium nitrate as a starting point, but while these results are interesting, later tests showing the amount of water absorbed in passing from an unsound t o a sound state shed a different light on the subject. The results of the tests are given in tabulated form. In Table I are also given the analyses and the fineness of the cements to which reference will be made later. I n one column are the results with lime sieved through a Ioo-mesh sieve and the cement a s received, and in another column the cement after passing through a 200- and lime through a Ioo-mesh. It is necessary t o explain that where “ Boiled 0 . .K.”
359
is used, it means the cement passed the test satisfactorily. To show degrees of soundness we use “Boiled 95 per cent.” to mean checked slightly; “Boiled 90 per cent.” to mean checked and softened slightly; “ Boiled 85 per cent.” t o mean softened considerably; and ‘‘ Boiled 80 per cent.” t o mean disintegrated entirely. These are only relative terms but they serve t o indicate different degrees of soundness, It will be noted in every case that the fineness of the cement played an important part, as when the particles coarser than zoo-mesh were removed, the ’
TABLE I. Cement through 200 mesh and CaO through 100-mesh percentages.
Cement as received and CaO through 100-mesh Fineness. Percentages. percentages.
-
NO.
7 -
6.0 2 mo.
1
..
c
4 mo.
1
.. 4
A
1 mo
3
B
1 yr....
yr
4 ,
yr
E
i
...{
2
{
97.5
85.5
96.0
79.0
,.,.( 9 3 . 0
..,.
1 yr....
I
28.0
90
10.0 3 0 . 0 12.5 33.0 15.0 35.0
95
12.5 15.0 17.5
I
6
I
Brand and 100- 200- (a) (6) CaO. age, mesh. mesh. CaO. Hfl. Boiled. (a) A m o . , . { 95.2 82.4 5 . 0 27.0 O.K.
so O.K. 95 90
(b)
Boiled.
12.0 15.0 20.0
34 35 39
O.K.
15.0 20.0
35 39
O.K.
95 80 90
20.0
38.0
80
25.0
45.0
O.K.
( C) 5.0 6.0 7.0
27.0 27.5 28.0
O.K. 95 90
8.0 10.0
30 33
90 80
6.0 7.0
27.0 27.0
O.K. 90
8.0 10.0
30 33
90 80
6.0 7.0
27.0 27.0
95
8.0 10.0
31 36
90 80
12.0 15.0
29.0 30.0
12.0 15.0 17.5
29 31 33
O.K.
75.4
96.8
33.0 35.0 36.0
O.K.
H20.1
78.6
O.K.
O.K. 95
95 80
Per cent. of lime added. ( b ) H 2 0 refers to the water used in making the pats. (c) Greater amounts of lime made the pats too soft to be boiled after 24 hours. (5)
-4NALYSES O F CEMENTS.
NO.
1.
2.
3.
Lossonignition‘.. Si02.... .. . . . . R& ............. CaO .... . . . . . . . MgO . ,
3.08 21.28 9.02 62.24 2.68 1.61
2.80 20.80 9.50 62.40 2.75 1.56
4.60 19.24 10.52 60.98 2.78 1.55
. .. . . ........ . . SOa......... . . . . .
4. 5.19 20.26 9.32 60.09 3.34 1.54
5.
6.
3.20 22.00 9.00 62.30 2.64 1.23
6.27 19.68 11.00 58.60 3.18 1.54
cement stood larger additions of lime. The inference is that the fineness of the cement, the fineness of the lime and its degree of fusion play an important part. Unfortunately, brands B, C, D and E were all over one year old when tested, but it will be noted that brand A, one month old, 8 5 % per cent. passing a zoo-mesh sieve, permitted I O per cent. greater addition of freshly calcined lime than the same brand per cent. two months old, 82.4 per cent. fine and z more than the same per cent. fine four months old, and that four coarser cements over a yearkold permitted the addition of less lime than the finer cement a t two and four months. Following this line and: believing that the-fineness of cement plays another -part, a sample of cement 1 The high loss on ignition in several of these brands is due to the long seasoning in bags,
360
T H E J O U R X A L OF I i V D L S T R I A L A N D EAVGI-VEERIIVG C H E i U I S T R Y .
80 per cent. passing a zoo-mesh sieve, which failed badly on the boiling test, was ground in a mortar to 8 5 per cent. fine and it showed improvement. I n another portion ground 90 per cent. fine, the improvement was more marked and in a portion ground 95 per cent. fine the boiling test was perfect. We repeated this on various samples of cement with similar results. Thinking that possibly the free lime became hydrated from the moisture in the air while being ground, similar experiments were made b y regrinding in a laboratory tube mill with a rubber gasket, so that the negligible amount of moisture in less than cu. f t . of air would have a negligible effect on hydrating the free lime in 1 5 lbs. of cement used in the test. Results were similar. Sieving t o various degrees of fineness in a small room where it was presumed the atmosphere was fairly dry, owing to the presence of a number of trays of cement and lime on which experiments were being made, 'gave similar results. Table I1 gives the results of the average of five different samples of cement in each case, each treated in the following manner: grinding in a mortar, grinding in a tube mill and sieving. I t will be noticed that the sample which was m.erely sieved to increase the fineness required a higher fineness as indicated by the zoo-mesh sieve than the reground samples. This, of course, was to be expected: TABLEII.-EFFECT
FINENESS ON
BOILINGTEST. B y reB y re- grinding Fineness grinding in tube regulated in mortar. , mill. by sieving. Roiling Boiling Boiling test. test. test. Per cent. Per cent. Per cent. SI per cent. passing 200 mesh, as received ...................... SO so 80 After bringing up to 85 per cent. fine 85 85 80 After fine .......................... 90 90 85 After bringing up to 95 per cent. . . . . . . . . . . . . . O.K. O.K. 90 After bringing up to 100 per cent. fine.. . . . . . . . . . . . . . . . . . . . . . . . . O.K. O.K. 95 OF
THE
I t might be argued that the free lime was largely in the particles removed b y the sieve, but this is offset by the tube mill experiments which were of exactly the same cement except as to fineness, as nothing was added or removed, and if the failure when 80 per cent. fine was due to free lime, the same free lime was present when i t was 95 per cent. fine, and if checking is due t o free lime, this free lime in the 95 per cent. fine was in a fine enough physical condition to become hydrated immediately on adding water when making the pat, or a t least t o become hydrated before the set had taken place. I n the coarser samples it is possible t h a t the free lime (if it is such) was present in the fused, glassy, coarser particles in such a way as t o prevent its hydrating before the set had taken place, and the expansion took place on boiling the pat after it had hardened. The inference here is plainly that the degree of fineness is an important factor. The foregoing experiments apparently lead t o the conclusion that a fair proportion of free lime
May, 1912
is not injurious to a Portland cement if its physical condition is such as to permit i t t o be hydrated before the set has taken place. Hydrated lime is added to Portland cement with beneficial effect in some cases and it matters not if the hydrated lime is added afterwards or free lime is in the cement when used, provided i t is in a condition to be readily hydrated, assuming that all other conditions of balancing the chemical composition and manufacturing details are properly regulated. I t is often noted that a sample of cement which fails t o pass the test one day may pass i t very satisfactorily several days later, and it has frequently been noted where the change from very poor to very good took place in 24 hours. I n some cases where the pats completely disintegrated on boiling, the cement became sound in three or four days, while with several brands samples have required thirty days or more I to become thoroughly seasoned. An effort was made to find what change took place in the ultimate chemical composition when a cement passed from an unsound condition to a sound one. If i t is due to a hydration of the lime by atmospheric moisture, then the amount of moisture absorbed would be a gage to the amount of free lime which had been hydrated, assuming (without any sound theoretical reason) that all the moisture taken u p would go to hydrating lime and none of i t enter into the complex reactions which take place when larger amounts of water are used. A number of experiments were made by determining the loss on ignition on unsound cements and making the same test after the cement became sound. A more correct method would have been t o absorb the water given off on ignition, as it is well known that cement absorbs carbon dioxide on seasoning and this also is given off on ignition, but a glance a t the results will show the relatively small total difference that even if we assume i t all to be water used in hydrating lime the amount of lime hydrated is comparatively small: 6 tests showed a difference of 0 . 4 5 per cent. in loss on ignition between the unsound and sound condition of the cement. 6 tests showed a difference of 0 . 7 5 per cent. 6 tests showed a difference of 0 . 9 0 per cent. Realizing the inaccuracy of the method, another set of experiments was conducted as follows: The unsound cement was placed on cover glasses and carefully weighed. This was placed on a rack under a bell-jar. A tray of water was placed below it and the whole set on an iron plate over a Bunsen burner. The bottom was sealed by piling finely pulverized flint (100-mesh and finer) around the lip of the jar so t h a t the air could not be renewed during the test. The temperature was raised to 150' F. and kept there from periods ranging from one t o twenty-six hours. A very great number of these tests were made and it was found that a t 26 hours' heating, three samples of different cements absorbed 1 . 2 7 per cent., 1 . 8 4 per cent. and z . 0 2 per cent. moisture. All of these, a? expected, passed the boiling test perfectly.
May,
1912
T H E J O U R N A L OF I S D V S T R I A L AND EiVGINEERI.YG C H E , W I S T R Y .
Repeated trials were made to determine the minimum amount of absorption necessary t o pass from an 8j per cent. failing cement to a perfect test: 3 tests showed 0 . I O per cent. was sufficient. . j tests showed 0 . 19 per cent. was sufficient. 3 tests showed 0 . 2 4 per cent. was sufficient. 3 tests showed 0 . 2 7 per cent. was sufficient. Whether these are the minimum amounts is open t o question, as a number of tests showed practically no absorption, yet it is apparent that moisture is essential to seasoning. Unsound samples kept in sealed jars a year still remained unsound, and samples heated to I j o o F. in dry air remained unsound, yet the above shows that a trifling amount of moisture produced the desired result. Moisture is essential and heat accelerates the action, as t o be expected. At lower temperatures it is necessary t o absorb considerably more water to get results than a t 150' F. Coarse cements required more water and the problem becomes more complicated where the degree of burning of the clinker, the fineness of the raw material, and other manufacturing factors are taken into account. Referring t o the above tests where the minimum water was absorbed, i t becomes apparent that in those cases, assuming all the moisture absorbed went t o hydrating free lime, the maximum amount of free lime exerting the injurious influence could not have been over 0.3 I per cent. Compare this with various amounts of free lime added, shown in Table I. I t may be possible t h a t the destructive expansion in the case of three cements requiring only 0 .I per cent. water in seasoning was caused by the hydration of its equivalent of 0 . 3 1 per cent. CaO, but it seems more probable that a physical explanation must also be considered in connection with it, for reasons briefly. stated as. follows : Lime calcined a t highest heat of a blast lamp is not injurious to soundness when added in reasonable amounts to Portland cements. Lime calcined a t the temperature attained in a rotary kiln does not become fused, hence when pulverized with the clinker should hydrate readily on the addition of water. I n burning of Portland cement clinker, small particles of free lime may be encased in the fused mass of silicates, aluminates, etc. If subsequent grinding reduces this glass to a finely enough divided state, this lime will hydrate on addition of water and have no injurious effects. If free lime is the cause of disintegration, i t seems that the state of fusion of the surrounding. glassy clinker is directly responsible for it. I t is claimed that over-limed cements are most apt to fail on this test. I t may be possible that owing to their being more difficultly fusible and requiring a higher temperature in the kiln, that a more glassy mixture envelops the free lime. This furnishes room for microscopic study. Under-burned cements may fail on this test. Free lime is given as the explanation of this. If free limes is the real cause, then the fact that as little as I/~,, of one per cent. of moisture will rectify many bad cases, leads to the belief that the phenomenon of seasoning is not so much one of hydrating free
361
lime as i t is a decrepitation process in which heat and moisture break up glassy particles, putting the lime in a physical condition to be hydrated, but not necessarily hydrating it. Acting on this idea, i t was believed that cements become finer with age. A number of tests seemed to confirm this, but the breaking of the small per cent. of coarser particles into finer ones takes place so slowly that differences of only one and two per cent. were obtained in sieving, which may not be sufficient to offset the possible personal error in the sieving process. To get a t the hydrating or decrepitation results, whichever i t may be, only the residues left on a zoo-mesh sieve were taken and sieved until no more would pass. These were exposed to air and moisture and again sieved each week. The following table shows the fineness of a sample of cement, part of which was exposed to air and part kept in a sealed jar for varying periods. Residues from the same cement were placed in an open-jar and starting with nothing that passed a zoo-mesh sieve, samples were sieved at various times, showing the per cent. passing the zoo-mesh after decrepitation had taken place. TABLEIII.-sHOWING
FINENESS ON -4
Sample exposed to the air. Per cent.
... ... . .. . . . . . . .. . . . . . . .. . . . . . .. . . . . . .. .. . . .
Starting point. . , After 14 days.. After 33 days.. After 49 days.. . After 9 1 days.. After 107 days..
84.6 85 .O 85,s 86 . O 86 .O 86.2
200--kfESH SIEVE.
Sample kept in a sealed jar. Per cent. 84.6 85.0 85.0 85.4 85.4 85.6
Sample of residues kept in an open jar. Per cent. 0.0 5.0 5.6 6.0 7.6 7.6
As was to be expected, the sample kept in air increased in fineness more rapidly than the one in the sealed jar, and while sieving tests are subject to some variation and inaccuracies, these samples were tested on the same sieve by the same man and by the same method, hence the personal error is probably about the same in each case. The results of the samples exposed to air might be explained by the theory of hydration of free lime, but those on sample kept in a sealed jar must be explained b y a physical theory of disintegration. SUMMARY.
While the work on this subject is by no means complete, i t represents the results of some hundreds of tests and a t the present time points to the following as worthy of consideration: I. Only a thin film on the coarser particle of cement enters into the reactions on the setting of the same. 2 . These coarser particles become finer with age. 3. The disintegration of these particles improves the soundness. 4. That if free lime is the cause of destructive expansion, three-tenths of one per cent. or less is sufficient to manifest itself. j . That cases were noted where practically no absorption of water took place, hence cannot be explained by the hydration of free lime. 6. Cement kept in a sealed jar becomes finer with age, thereby indicating a mechanical decrepitation,
362
T H E JOURNAL OF I N D U S T R I A L A N D EiYGIiVEERING C H E J I I S T R Y .
which, as tests show, becomes more pronounced in the presence of moisture. 7 . Carbon dioxide from the air, while i t may aid in seasoning, is not essential, as shown b y several series of tests. 8. That increasing the fineness of any particular sample of unsound cement improves its quality in this respect. 9. With the toregoing data, it appears that the theory of free lime as the sole cause of unsoundness is not well founded. I O . Expansion may be due principally to physical strains in the coarser particles. 1 1 . Any theory of expansion on a purely physical basis seems inadequate, inasmuch as pats may frequently remain in boiling water several hours or more before destructive expansion becomes manifest. If i t were merely the different rates of expansion, one would expect the rupture to take place shortly after the boiling begins. 12. With the data in hand a t present, a possible theory presents itself as follows: Seasoning of a n unsound cement consists of decrepitation of the coarser particles, caused by or accelerated b y heat and moisture, and that the moisture need not be in sufficient quantities to hydrate the free lime, but merely t o render the coarser particles sufficiently fine t o permit hydration before the set has taken place. 13. Le Chatelier claimed tri-calcium silicate to be the essential constituent of Portland cement and free lime as the agent destructive to soundness. Unger obtained vitrified tri-calcium silicate which disintegrated (seasoned), yet hardened well, while a fused tri-calcium silcate was unsound. This would indicate t h a t if free lime is present in Portland cement, it is in the glassy unground particles. 14. 0. Schott denied the existence of tricalcium silicates, tri-calcium aluminates, and tri-calcium ferrites in Portland cement, yet fused mixtures of these compositions showed the characteristic destructive phenomena found in unsound cements and which he attributed t o free lime. 1 5 . Michaelis claimed burning a t too high temperatures produced unsound cements. Microscopic examinations of clinkers show a different physical structure of that from a rotary kiln and that burned at lower temperatures. The contradictory results obtained by investigators, both of synthetical mixtures and actual clinker, seem to be due t o different degrees of fusion. 16. Shepherd and Rankin have demonstrated the existence of calcium tri-silicate, but the real cause of expansion is yet not conclusively proven. 17. What is needed now, in addition t o the excellent work of the authorities herein mentioned and others unmentioned, is an investigation which will link together our knowledge on purely theoretical constitution of Portland cement clinker and the chemical and physical constitution of a commercial Portland cement, and further show what changes take place in passing from a n unsound to a sound state.
May, 1 9 1 2
THE CHEMISTRY O F ANAESTHETICS, 1V: CHLOROFORM. By CHARLESRASKERVILLEA S D W. A. HAMOR. Received December 13, 1911.
(Continued from the A p r i l No.) 2. T H E
CHAKGES
UNDERGOES
WHICH
WHEN
A
AN.IESTHETIC CURRENT
OF
CHLOROFORM OXYGEN
IS
COKDUCTED THROUGH IT.
Among the anaesthetic mixtures, the combination of Chloroform vapor with oxygen was used shortly after the introduction of chloroform as an anaesthetic, and it has recently been reintroduced into practice by Neudorfer, Kreutzmann, and others. I t is stated b y anaesthetists that oxygen does not antagonize the action of chloroform on the heart or nerve centers, but that it protects the patient from the dangers which result when chloroform is administered while his blood is in a condition of undue venosity' and that it prevents any intercurrent asphyxial condition. Gwathmey has stated positively' that oxygen increases the value of all inhalation anaesthetics as regards life. I t has been maintained, however, that chloroform undergoes alteration in this procedure. Falks attempted t o demonstrate that the passage of oxygen through chloroform4 produces chemical changes in the anaesthetic. He reported that after the passage of oxygen for 2 0 minutes, changes could be recognized in the residual chloroform, in some cases hydrochloric acid and in others an acid with reducing properties5 having been recognized. The quantities produced were found t o be greater the higher the temperature and degree oE illumination. This work is partially contradicted b y the clinical results obtained b y anaesthetists, and b y the observations of Willcox and Collingwood6 on the administration of oxygen bubbled through absolute alcohol.7 In order to determine whether the passage of a current of oxygen through anaesthetic chloroform results in the decomposition of the anaesthetic, 4 ounces of chloroform containing 0 . 7 0 per cent. b y volume of absolute alcohol and 0.035 per cent. by volume of water, but otherwise pure, were placed in the chloroform container of the Gwathmey anaesthetic apparatus (Fig. I ) , and a slow stream of oxygen was conducted through the chloroform for ten and one-half hours, a t the end of which time 3 / / , ounce of chloroform remained. The experiment was conducted under the usual conditions obtaining during administration by the Gwathmey method, b y electric light and a t an average temperature of z o o C. The residue, upon examination, gave the following results: Buxton, Araesthefics, 4th ed., p. 299. Medical Record, October 8th, 1910, p. 616. "On Chloroform-Oxygen Narcosis," see also Ziegncr, Miinch. Med. Wochenschr., 67, 2585 : and Guarini, Sci. Am.. 90, 24. 3 Deut. Med. Woch., 1902, 862. 4 The purity of this was not described, but it was evidently of the grade specified by the German Pharmacopoeia. I Acetic acid, resulting from the oxidation of the alcohol in the chloroform used (?). 8 Brit. Med. J . , Nov. 5th. 1910. 7 Willcox and Collingwood stated that the administration of oxygen bubbled through absolute alcohol is a marked cardiac stimulant. It is especially important to note that they found the administration pleasant and non-irritating to the patient-that it causes no ill-effects t o the lungs or bodily system. 1
2
.