37 N.-On some New Derivatives of Coamarin. By W. H. PERKIN, F.R.S. INMay last I had the honour of laying before the Societya short account of some derivatives of this interesting product. Since thar time I have procured further information upon the sub- stances I then described and have also obtained a number of new derivatives of coumarin. I now beg leave to bring an account of my results before the Society. Dibromide of Cozimarin. When preparing this product I originally mixed a solution of coumarin in disulphide of carbon with a similar solution of bro-mine using these substances in theoretical proportions The resulting solution mas then allowed to evaporate spontaneously and the product purified. By this method I obtained a very poor yield of the dibromide.I have since found that if the solution of bromine and coumarin be kept for twelve hours or more previous to the evaporation of the disulphide a much larger yield of product is obtained 5 grms. of coumarin gener- ally yielding from '7& to 8 grms. of dibromide. Dibromide of coumarin when mixed with an alcoholic solu- tion of potassic iodide becomes brown and on evaporation deposits metallic-looking needles apparently consisting of a mixture of iodine and coumarin crystallised together. u. Brornocounaarin. Upon referring to the previous method given for the pre paration of this body it will be found that it wa8 obtainedfiom the mother-liquors of dibromocoumarh. If however a mixture of bromine and coumaiin (mixed with disulphide of carbon) in the proportion of two parts of the former to one of the latter be heated in a sealed tube to about 200' C.for three or four hours the resulting product will consist almost entirely of monobromocoumarh. In this reaction dibromocoumarin is undoubtedly first pro- duced but at the high temperature employed it is decomposed VOL. xxxv. E PERKIN ON SOME NEW by the hydrobromic acid produced losing half of its bromine thus-C,H,Br,O + HBr = CSH,BrO + BrBr. Dibromocoumarin. Bromocoumarin. The bromocoumarin obtained by this process sometimes crystallises from alcohol in long slender needles ; these how- ever on standing in the mother-liquor for a few days become short and hard like those described in my previous paper; in fact so different was the appearance of this body when freshly crystallised from that which 1had previously obtained that I sub-jected it to analysis.The following are the results obtained :-I. -4440 of substance gave -3710 of AgBr. 11. -3436 of substance gave *5986of CO, and *0702of H,O. Theory. Experiment. # h -. I. 11. C .... 108 48.00 -47.51 H .... 5 2.22 -2.2 7 Br .... 80 35.56 35-56 -14-22 -0 .... 32 -225 100*00 I have however found a still more simple process for the preparation of bromocoumarin than the above. It consists in the decomposition of the dibrornide of coumarin wit11 alkalies the following reaction taking place :-CI,H,O,,Br + KHO = C9H,Br0 + KBr + H,O. Dibromide of coumarin. Bromocoumarin.When this process is employed the dibromide of coumarin is powdered and suspended in alcohol and rather more than enough alcoholic potash is added than is required to complete the above reaction. The decomposition takes place rapidly the mixture becoming quite hot and of a pale yellow colour. After it has stood for some time water is gradually added until it ceases to throw down any more bromocoumarin. When sufficiently diluted the product is collected upon a filter washed DERIVATIVES OF COUMARIN. dried and crystallised from spirit. In this way it is generally obtained in transparent needles about half-an-inch in length. An analysis of this product gave the following numbers :--4102 of substance gave *7242 of CO, and *0872of H,O.Theory. h - Exp. c .......... 108 48.00 48.15 H .......... 5 2.22 2-36 Br .......... 80 35.56 0 .......... 32 14-22 225 100*00 The alkaline filtrate from the above contains a quantity of this substance in solution which may be precipitated with acid and if a large excess of alkali has been used most of the bro- mocoumarin will be found in solution ; it is however apparently not usually so pure as that precipitated with water and gener- ally forms smaller crystals. From the difference in the appearance of the products ob-tained by the above processes I was at first inclined to think they were isomeric forms of bromocoumarin but as the melting points are nearly identical as are also the products of decom- position I am induced to believe that these variations are due to the presence of small amounta of impurities.a Bromocoumarin when left in contact with cold alcoholic ammonia decomposes with formation of ammonic bromide and a non-crystalline sticky mass easily soluble in water. Heated with potassic hydrate it yields potassic bromide and a new acid. It decomposes when heated with potassic cyanide and alcohol in a sealed tube forming a brown solution from which water throws clown a drab coloured amorphous precipitate. When it is heated with alcohol to 200' C. in a sealed tube for five or six hours slight decomposition takes place with formation of hydrobromic acid A similar change takes place if water be used instead of alcohol. a. Uibromocournarin. The method I previously gave for the preparation of this body consisted in heating in a sealed tube to 140' C.a mixture E2 PEREIN ON SOME NEW of one part of coumarin two parts of bromine and four or five parts of disulphide of carbon. I have since found however that this process is greatly improved by the addition of iodine to the mixture as it is then only necessary to heat the sealed tube for four or five hours in a bath of salt and water to com-plete the reaction. The procinct is freed from disulphide of carbon by evaporation and from iodine by means of a solution of potassic iodide and finally purified by two or three crystalli- sations fi-om alcohol. A specimen prepared in this manner gave the following numbers :--2584 of substance gave *3330of CO and -0371 of H,O.Theory. Exp. C .......... 108 ~ 35.53 35.15 H ......... 4 1.31 1-59 Br .......... 160 52-63 - 0 .......... 32 10.53 L__- 304 100*00 The melting point of this substance is 183' C. and not 174' C. as I previously gave it. It is however not so definite as might be desired as it generally shows signs of fusion below 183O C. u Dibromocoumarin is easily decomposed by boiling alcoholic or aqueous potassic hydrate with formation of potassic bromide and the gotassic salt of a new acid. When suspended in cold alcoholic ammonia 01 dibromocou-marin becomes of a pale yellow colour and gradually dissolves with formation of animonic bromide ; on standing a Crystalline product is deposited. It is volatile when heated and evolves ammonia when treated with potash.The resulting alkaline solution does not deposit any product when acidified. It is soluble in water and crystallises in beautiful white needles. The alcoholic solution upon evaporation yields a second crystalline body soluble in water. It cannot be volatilized without undergoing decomposition. A syrupy body is alao produced. I hope to return to the examination of these products at a future period. DERIVATIVES OF COUMARIN. p Bromocoumarin. It will be remembered that the sodium derivative of the hydride of salicyl when digested with acetic anhydride yields ordinary coumarin. It therefore appeared to be of interest to treat the sodium compound of brominated hydride of salicyl in a similar manner to see whether a brominated coumarin could be obtained.The hydride of sodium-bromosalicyl when submitted to the action of acetic anhydride rapidly changes colour and dissolves the mixture becoming quite hot. After the reaction has mode- rated the product is well boiled for ten or fifteen minutes and then poured into water in which it sinks as a heavy oil acetate of sodium dissolving. On distilling this oil acetic an- hydride and acetic acid first come over then a quantity of oily product which rapidly solidifies ; the first half of this contains a large quantity of the hydride of bromosalicyl the remaining portion however when crystallised from alcohol two or three times yields col ourless flat prisms apparently rhombic. This substance gave the following numbers on analysis :-I.*5170of substance gave -9071 of CO, and -1036 of H,O. 11. 03576of substance gave -6276 of CO, and *0735of H,O. The numbers give percentages agreeing with those required by the formula- C9H5Br02 as the following comparisons will show :-Theory. Experiment. y-'I. 'Y 11. C .... 108 48.00 47.84 47-86 H .... 5 2-22 2-22 2.28 Br .... 80 35-56 c 0 .... 32 14.22 -225 100~00 This substance is therefore monobromoco~mar~n.It is PERKIN ON SOME NEW moderately soluble iii spirit and when fused has an odour some- what similar to that of coumarin. It greatly differs in properties from the bromocoumarin previously described. Its melting point is 160' C. or 50 degrees higher and when boiled with alcoholic or aqueous potassic hydrate it does not decompose with formation of potassic bromide but simply dissolves like ordinary coumarin I have therefore termed it p bromocou-marin.p Dibromocoumarin. On treating the hydride of sodium-dzlxomosalicyl with acetic anhydride in exactly the same manner as for the preparation of bromocouniarin a beautifully crystalline product is ob-tained. It gave the following numbers on analysis :-I =3639of substance gave- 04714of CO and -0463 of H,O 11. 02080of substance gave- 9623 of GO2 and 00266of H20. These numbers give per centages agreeing with those re- quired by the formula- C9H41Jr202 as will be seen by the following comparison :-Theory. Experimeiit. h v 11.C .... 108 35.53 35.33 35.41 H .... 4 1.31 1.41 1-46 Br .... 160 52.63 -0 .... -32 10.53 -304 100~@0 The substance is therefore dibromocoumarin. It is rather difficultly soluble in alcohol. It crystallises in short hard needles. It is not the same body as that obtained by acting on coumarin with bromine and iodine. It melts at 176' C. and is not decomposed by boiling with a solution of potassic hydrate. I have therefore designated it as dibromocoumarin. DERIVATIVES OF COUMARIN. Dichloride of Counzarin. A solution of coumarin in chloroform absorbs chlorine gas very minute quantities of hydrochloric acid being formed. On allowing the solution to evaporate spontaneously after the chlorine has been passed through it for an hour or two a syrupy product is obtained very like new honey.This is the dichloride of coumarin. I have not analysed it as there is no way of judging of its purity; but from its products of decom-position there can be no doubt that it possesses the formula- C&4p,,C1,. On keeping it appears to decompose; when heated it gives off hydrochloric acid and on distillation is converted into chlorocoumarin. With a1 coholic potash it decomposes in the same way as the dibromide. a. Chlorocournarin. When a mixture of one part of coumarin and three parts of pentachloride of phosphorus is heated in a retort placed in an oil-bath the two bodies slowly react upon each other as the temperature rises and when the oil has reached about 200°C. the product becomes a dark brown liquid (at a few degrees higher it is converted into a carbonaceous mass) ; during this reaction a volatile liquid consisting chiefly of terchloride of phosphorus distils over.The contents of the retort after treatment with water become a pasty mass of crystals which is first purified by distillation and then by several crystallisations fiom alcohol. The substance gave the following numbers on analysis :-1. 03904of substance gave- 3546 of CO and *I050of H,O. 11. 04006of substance gave- -8760 of CO and 01047of H,O. 111. 03364of substance gave- *2703of AgC1. These numbers give percentages agreeing with those re- quired by the formula- C,H,ClO,. PERKIN ON SOME hiW The following is a comparison of the theoretical and experi- mental numbers- Theory.Experiment. F -I. 11. 111. -C . . .. 108.0 59-83 59.70 59-64 H .... 5.0 2.77 2-98 2.90 -C1 .... 35.5 19.67 -19-88 0 .... 32.0 17.73 -I --c-180.5 100*00 From these results it is evident that the pentachloride of phosphorus has simply given up chlorine to the coumarin. It is very curious that if half the amount of pentachloride of phosphorus mentioned above be employed the product of the reaction suddenly carbonizes sometimes when the oil-bath is at a temperature as low as 150'C. I have had this happen twice whereas with the excess of pentachloride the product may be heated in an oil-bath standing at 200' C. Chlorocoumarin may be easily prepared by treating the dichloride of coumarin with alcoholic potash in exactly the same manner as described for the preparation of ct bromocou-rnarin the reaction being analogous- CgH60,C12 + KHO = C,H5C10 + KC1 + H,O.Dichloride of coumarin. Chlorocoumarin. Obtained by either of the above processes this substance forms flat needles about half an inch long. It is moderately soluble in alcohol and slightly so in hot water from which it crystallises on cooling. It melts at 122°-1230 C. and when heated emits an agree- able aromatic odour. When boiled with alcoholic potassic hydrate it decomposes with formation of potassic chloride and forms the same product as a bromocoumarin when treated in the same manner. From this description ct chlorocoumarin is evidently a sub-stance difl'erent fiom that obtained by Dr.Basecke his product is in fdct the 6 chlorocoumarin corresponding to the bromo-coumarin and was obtained fiom the hydride of chlorosalicyl. DERIVATIVES OF COUMARIN. Tetrachlorocoumarin. Chloriiie gas when passed through a solution of coilmarin and iodine in tetrachloride of carbon is rapidly absorbed hydrochloric acid being evolved; if the gas be passed for two or three hours a quantityof a reddish body separates on evaporat- ing the product so as to separate the tetrachloride of carbon an oily residue is obtained the red substance having fused with the impurities. On mixing this substance with alcohol it soon becomes a white paste. This is then pressed in a small linen bag when a white product is obtained which is further purified by being several times crystallised fiom spirit.It gave the following numbers on analysis :--3877 of substance gave- 05309of C02 and *0355of H,O. These numbers give per centages approximating to those required by the formula- C,H2c1402 arJ will be seen from the following comparison :-Theory. EX~. -C9 .......... 108 38.03 37.32 H .......... 2 0.70 1-02 C14 .......... 142 50.00 - 0 .......... 32 11-27 - -284 100-00 It is therefore a tetrachlorocoumarin. This siibstance melts at 144'-145' @. It is difficultly soluble in spirit from which it crystallises in small white needles. When heated with alcoholic potassic hydrate it decomposes with formation of potassic chloride and the salt of a new acid not yet examined.Coumarilic Acid. I have already mentioned that tc bromocoumarin when boiled with a solution of potassic hydrate decomposes yielding potassic bromide and the aalt of a new acid. To prepare this acid in a PERKIN ON SOME NEW pure atate I have found it beat to proceed in the following manner A quantity of pure tc brornocoumarin is mixed with an excess of the ordinary solution of potassic hydrate. On heating this mixture the bromocouniariii gradually dissolves but on reaching the boiling-point the solution rapidly becomes a pasty crystalline mass. Sufficient water is then added to dissolve this and the boiling continued for about an hour. The solution on cooling deposits the potassic salt of the new acid in small needles.If any quantity of this salt remains in the mother- liquor the addition of potassic hydrate will precipitate it. The crystalline salt thus obtained is collected upon a linen filter and well pressed to remove as much of the alkaline-mother liquor as possible. It is then crystallised once or twice from alcohol and dried. To obtain the acid from this salt it is dissolved in water and hydrochloric acid is added in excess. The new acid is then thrown down as a snow white crystalline precipitate which is purified first by washing with water on a filter and then by cry stallisation from boiling water. It gave the following numbers on analysis :-I. 9112 of substance gave- 05146of CO and -0734 of H,O. 11. 9561 of substance gave- 06266of CO and -0846 of H26.These percentages give numbers agreeing with bhose re-quired by the formula- %HBOS as the following compa.rison will show :-Theory. Experiment. / h \ r---\ I. 11. C .... 108 66.67 G 6-45 66-72 H .... 6 3.70 3.8 6 3.67 I 0 .... -48 29.63 -162 100*00 This acid which I have proposed to call coummdic acid melts at 192O-193OC. It distils without leaving any residue but decomposes partially with formation of an oily body smelling DERIVATIVES OF COUMARIN. like naphthaline. When gently heated it sublimes. The vapour is very suffocatinq. It is excessively soluble in alcohol but difficultly so in chloroform and disulphide of carbon. It is moderately soluble in boiling water from which it crysta11' ises on cooling in beautfful long needles like benzoic or cinnandc acid Its aqueous solution has a bitter taste.It is not decomposed when heated with potassic hydrate to 180O C. for an hour O! chlorocoumarin may be employed instead of cc bromocou-marin in the preparation of this acid. It is monobasic. dmrnonic Coumarilate.-This salt is easily obtained by dissolv- ing cournarilic acid in aqueous ammonia and evaporating in vacuo over sulphuric acid. It is a beautiful salt crystallising in flat prisms radiating from a commoii centre. It is easily soluble in water. Xodic Coumarilak.-Coumarilic acid dissolves in a solution of sodic carbonate with effervescence forming a salt which crys- tallises in transparent rectangular tables. It is easily soluble in water.Potassic Cozcrna&late.-The preparation of this salt has already been given. It is difficultly soluble in cold alcohol but moderately soluble in this menstruum when boiling ; nearly insoluble in solutions of potassic hydrate of about 20 per cent. It crystallises in pritJms half an inch long which break up into sinall plates. Its taste is bitter. A determination of the potas- sium in this salt gave the following numbers :-*1781of salt gave- -0759 of K,SO = 19-07per cent. of potassium. The formula C9H,K0 requires 19.5 per cent. of potassium. CaZcic Coumarilate is thrown down as a precipitate on addition of calcic chloride to potassic coumarilate. It is crys-talline and difficultly soluble in water. Baric Cournarilate is somewhat like the calcic compound.Afyentic Coumarilate. -On adding argentic nitrate to a solution of potassic coumarilate this salt is thrown down as a white curdy precipitate slightly soluble in water. The follow-ing determinations were made of the silver in this salt :-I. -1118of substance gave- *0443silver = 39.62 per cent. PERKIN ON SOME NEW 11. 01608of substance gave- -0648 of silver = 40.29. The formula C9H,Ag0 requires 40.15. Analysis I1 was made from coumarilic acid prepared from u chlorocoumarin 1 from u bromocoumarin. Plumbic Coumarilate is a white curdy precipitate. Mercurous Cownarilate is also a white precipitate. Ferric Coumarilate is a pale brown precipitate. BPQomocoumarilicAcid. This acid is produced in the same manner as coumarilic acid but substituting u dibromocoumarin for bromocoumarin.LY The acid is also purified by crystallisation from spirit and water instead of water. It gave on analysis the following numbers :-I. 02504of substance gave- -4109 of CO and *0489of H,O. 11. 93260 of substance gave- -5357 of CO and *0658of H,O. Thefire numbers give percentages agreeing with the formula- C9W5Br0, a8 the following comparison will show Theory. Experiment. 7-\ I. 11. C .... 108 44.81 44.75 44.82 H .... 5 2.08 2.17 2.24 Br .... 80 33-19 -0 .... 48 --19.92 241 100.00 I therefore propose to call this acid brornocoumnrilic acid; its formation may be expressed thus :-C9H,Rr,0 + H,O = C,H,BrO + HBr. Dibromocoumarin.Bromocoumarilic acid. Bromocoumarilic acid melts above 250' C It crystallises DERIVATIVES OF CUUMARIN. from a mixture of spirit and water in needles. It is very soluble in alcohol and but little so in water. When heated to 180' C. with potassic hydrate it decomposes and becomes brown potassic bromide being formed. It has a bitter taste andis monobasic. Ammonic Bromocozcmarilate is easily soluble in water and crystallises in needles. Sodic Bromocoumarilate is easily soluble in water but does iiot produce good crystals. Potassic Bromocoumari1ate.-This salt is prepared like the corresponding coumarilate. It is easily soluble in water but is almost wholly thrown down from its aqueous solution by strong solutions of potassic hydrate.It crystallises from alcohol in beautiful long needles. Determinations of the potassium in the salt gave the following numbers :-I. 03215 of substance gave- -1026 of K,SO =.14*30per cent. of K. 11. el897 of substance gave- *0600of K,SO = 14-17 per cent. of l<. The formula C9H,BrK0 requires 14-00per cent. of potassium. Bark Bromocoumarilate may be obtained by the addition of a solution of baric chloride to the above potassic salt as a crys- talline precipitate difficultly soluble in water. Argentic Bromocoumarilate is a white precipitate. Plumbic Bromocoumarilate is a white precipitate. Su@liocoumarilic Acid. On digesting a mixture of about one part of coumarin and five parts of fuming sulphuric acid in the water-bath for an ho& or two the product will be found to dissolve perfectly in water.This when diluted and neutralized with baric carbonate yields besides baric sulphate a soluble salt which when evaporated crystallises in tufts of transparent prisms. This is rendered pure by recrystallisation from water. To obtain the new acid from this salt it is dissolved in water and sulphuric acid added in exactly the quantity necessary to precipitate all the barium as sulphate. The solution is then filtered and evaporated first over the water-bath and then in vacuo. In this way the acid is usually obtained in needies easily soluble in water. It has not been burnt but from the PERKIN ON SOME NEW analysis of its salts its formula when anhydrous is C,H,O,SO,.Determinations of its water of crystallisation gave the following numbers :-I. 09330of crystallised acid lost 01305of water at 100°C. = 13.99 per cent. 11. 05635 of crystallised acid lost -0783 of water at 100' C. = 13.9 per cent. Ti1 one experiment I obtained this acid in octahedia which were very brilliant whilst in the mother-liquor but became opaque as soon as exposed to the air. A determination of the water of crystallisation gave the following results :-*3792of' substance lost 00527 of water at 100OC. = 13.90 per cent. This preparation when recrystallised was deposited on the palm of needles. It is therefore probable that they originally crystallised with more wat'*erthan is contained in the needles. These numbers correspond with those required by the formula-C9H60,S0,2H,0 which theoretically contains 13.74 per cent.of H20. The crystallised acid does not lose any more weight at 200' C. than at 100"C. This acid which I propose to call sulphocoumarilic acid possesses a strongly acid and bitter taste; it is very soluble in water and may be evaporated to a thin syrup before it shows any signs of crystallisation. It is monobasic. When its salts are mixed with an excess of alkali a yellow solution is obtained similar in appearance to that produced by dissolving coumarin in alkali. Ammonic Sull3J~ocournnrilate.-This salt is obtained by eva-porating a solution of the above acid to which an excess of ammonia has been added. It crystallises in satiny needles very soluble in water and also soluble in alcohol.Sodic SuZpJ2ocoumariZate.-On adding a solution of sodic carbo- nate to n solution of sulphocoumarilic acid effervescence takes place and on evaporating the resulting solution when neutral beautiful crystals are obtained very transparent and derived fkom the rhombic octahedron. It is easily soluble in water but nearly insoluble in alcohol. DERIVATIVES OF COUJIARIN. Potassic Sulphocoumarilate crystallises in flat prisins easily soluble in water insoluble or nearly so in alcohol. Ba& Xu113hocournarilate.-The preparation of this salt has already been given. It contains a considerable quantity of water of crystallisation the crystals efflorescing rapid17 in vacuo. The formula is-C,,Iq,,O43a"~S0,5H,O as the following results will diow :-I.1.0232 of crystallised salt lost in vacuo-*lo67 of H,O = 1043 per cent. and from vacuo to 18OOC. *0279= 2-90 per cent. IT. 07420of crystallised salt lost in vacuo-00777 of H,O = 10.50 per cent. and from vacuo to 180 00198= 2.66 per cent. These numbers show that this salt loses four molecules of water in vacuo the fifth being given off entirely at 18OOC. Theory. Experiment. 11.-' I. 1 4 mol. ...... 10.63 10.43 10.50 1mol. ...... 2-66 2.90 2.66 Determinations of' baiium in the anhydrous salt gave the following numbers :-I. 02456of substance gave- -0969 of baric siilphate = 23-20 per cent. of barium. 11. *1192 of substance gave- -0474 of bark sulphate = 23.38 per cent. of barium. 111.*0994of substance gave- *0393of baric sulphate = 23.25 per cent. of barium. The formula C,,H,,04Ba"2S0 requires 23.34 per cent. of barium. Combustion of barium salt containing one molecule of water. I. -2911 of substance gave--3818 of CO, and -0570 of H,O. PERKIN ON SOME NEW 11. *3850 of substance gave- 05010of CO, and -0787 of H,O. These give numbers agreeing with the formula -C,,H,,O,Ba”2 SO,H,O ae the following comparisons will show :-Theory. Experiment. c f A % I. 11. c,,.... 216 35.70 35-77 35.70 Hi,. .. 12 1-98 2.17 1-98 O,,.... 176 29.09 I -Ba .... 137 22-65 s .... 64 10.58 -605 1oo*oo Strontic Xulphocoumal.ilate.-This salt is easily soluble in water and crystallises very nicely. It contains water of crystallisation becoming opaque at 100” C.the last quantities of water not being expelled until a much higher temperature is used. A determination of the strontium in the anhydrous salt gave the following numbers :--1698 of substance gave- =0574of strontic sulphate = 16-15 per cent. strontium. The formula Cl,Hlo0,Sr”2S0 requiring 16.27 per cent. -0769 of a second specimen dried at about 110’ C. gave -0253 of strontic sulphate = 15.69 per cent. of strontium. The formula C,,HlO0,Sr”2SO, H,O requiring 15.75 per cent. Disulphocoumarilic Acid. On heating a mixture of about eight parts of fuming sul- phuric acid and one part of coiimarin to a temperature of 150” or 160’ C. for an hour or two the product contains two sulpho-acids viz.sulphocoumarilic and disulphocoumarilic acids. On treating this with baric carbonate in the usual manner the solution contains the barium salts of these two acids. To separate them the solution is evaporated to dryness and treated DERIVATIVES OF COUMARIN. with warm water which removes the baric sulphocoumarilate leaving the less soluble salt of the new acid behind. This is purified by crystallisation from boiling water in which it is not very soluble. Thus obtained it is a white slightly crystalline product nearly insoluble in cold water. Dried at 100’ C. it gave on analysis the following numbers :-03258of substance gave- 01648of baric sulphate = 29.76 per cent. of barium. The formula CSH,02Ba”2S0,,H,0 requires 29.85 per cent.of barium. Determinations of water of cry stallisation- I. 1.4226 of salt dried at 100’ C. lost 00486of H,O at 180’ C. = 3.42 per cent. 11. *5612of salt dried at 100”C. lost *0235of H,O at 180’ C. = 4.18 per cent. The above formula require8 3-92 per cent. Determination of barium in anhydrous salt- 01904of substance gave *0994 of baric snlphate = 30.7 per cent. The formula CSH,0,Ba”2S0 requires 31.06 per cent. From what has been said of the bromine and chlorine deriva- tives of coumarin it is evident that the bromine or chlorine introduced must be related to Merent carbon groups according as the product belongs to what has been termed the u or p derivative. In the case of the latter it is evident that it is contained in the C group as these bodies are produced from the brominated hydride of salicyl; those employed in this paper being represented thus :-Hydride of bromosalicyl.Hydride of dibromosalicyl. The bromine in the bromocoumarin obtained &omt)hese bodies by means of acetic anhydride is not attacked by boiling with potash. In the alpha series the bromine or chlorine must be connected VOL. XXIV. F PEREIN ON SOME NEW with the carbon of what may perhaps be called the acetic residue. These are easily decomposed with caustic alkali. I have endeavoured to produce a bromocournarin of this class by employing bromacetic anhydride instead of acetic an- hydride in the preparation of coumarin from the hydride of salicyl but have not yet succeeded. In OL dibromocouinarin one half of the bromine appears to be in union with the C group the other being connected with the acetic residue as only one atom of bromine is removed by boiling with potassic hydrate.a Bromo-and clilorocoumariiis from their decomposition by caustic alkali and production of couniarilic acid would appear to to be bromide and chloricl\ of :inacid radical (coumaryl). If it mere so however these substances should give with ammonia an amide which wodd decompose on boiling with alkali yielding ammonia and coumarilic acid ; but although the substances are easily acted upon by annionia the product when boiled with potassic hydrate does not yield this acid. It is therefore possible that coumarilic acid is not a true acid containing the group COHO but is only coumarin with one of hydrogen replaced by hydroxyl unless some molecukr change takes place during its formation ; and it i8 perliaps worth remarking that coumarin and its bromine chlorine and sulphuric derivatives give yellow solutions with caustic allrali whereas coumarilic acid forms colourless ones ;this may perhaps point to a change in struc- ture.I hope however to replace the hydroxyl in coumarilic acid by bromine and see if the resulting conipound is ordinary u bromocoumarin. Fittig and others consider coiimarin as the anhydride or glycollide of oxycinnamic rtcicl and it appeared to me probable that if it weie so it should yield nith phosphoric chloride the chloride of clilaocinnamyl. I have not however as yet suc- ceeded in obtaining this body clilorocoumarin being thc principal product found.I must say that if coumarin be a glycollide it is a very 8110-malous one in its properties as it may be kept d&sol& in caustic alkali for months without formation of any appreciable quantity of comnaric acid; neither does it form an arnidc when heated alone in gmeous ammonia or in a sealed tube Tvith alcoliolic ammonia ;whereas xalicylide the glycollicle of salicylic acid is converted into a salicylate as quickly as it dissolves in DERIVATIVES OF COUMARIN. an alkali and from the description given it would appear that it cannot exist in solution in alkali. Coumarilic acid differs in composition from coumaric acid by H, and from inelilotic and phloretic acids by H, so that we have the following series :-CgH,03 coumarilic.C,H,O coumaric. C9H,,0 melilotic and phloretic. These acids are however very different in properties for example phloretic and melil otic acids distil with formation of their anhydrides and couinarilic acid distils with only partial de- composition whereas the intermediate compound couniaric acid when heated forms a brown resinous mass and when distilled is entirely split LIP; if coumarin were its anhydride we should expect it to be produced especially as it is very stable and easily volatilized. The following is a list of the new products described in this paper :-Dibromide of coumariri .......... 'gH6 02,Hr2* Dichloride of coumarin .......... CgH60,Cl,.a. Bromocoumarin .............. CgH,BrO,. p. Bromocoumarin .............. C,H,BrO,. a. Chlorocoumarin .............. CgH5C10,. p. C~L lorocozcrnarin .............. C9H5Cl0 (Basecke). a. Dibromocoumarin ............ CgH,Br20,. ,8. Dibromocoumarin ........... C,H,Br,O,. Tetrschlor ocoumarin ............ CgH,C140,. Coumarilic acid ................ C,Hs03* Bromocoumarilic acid ............ CgH5Br0,. Sulphocoumarilic acid. ........... C,H602S0,. Disulphocoumarilic acid .......... C9H60,2S03.