209 



In the \v<>rk of the Bureau of Chemistry for the pa.-t year nr more much work lias 

 been done to perfect the Allen-Mar<|iuirdt method, and it will be discued some 

 what in detail, as some of the moditicatioiis deviled improve the method and have 

 not been published, but are of great importance in obtaining accurate results. 



The method as used at the present time is the same as is given in Bulletin 107, 

 Ke\ i.-cd. but it has been found necessary to have the proof of the sample under exami- 

 nation not much above 100 in order that the volume when made up to 1.12 specific 

 gravity will not be too great for the separatory funnels used. In the analysis of hi^h 

 proof spirits, therefore, 50 cc of the sample are used for analysis and 50 cc of water 

 added, making the product approximately 100 proof. 



( )ne point which is extremely important is that the carbon tetrachlorid used must 

 be of the hi.irhe-t purity. \Ve have found that most of the ('. 1'. carbon tetrarhlorid 

 on the market is entirely un.-ati-factory for this determination until it has been puri- 

 fied by oxidation with bichromate and sulphuric acid, as called for in the present 

 provisional method. In the projKT control of this purification a renewal of the 

 bichromate and sulphuric acid mixture is necessary and often makes the procew a 

 lengthy one \\here the carbon tetrachlorid is very impure. For this reason the fol- 

 lowing new method, devised by A. M. Breckler, is used: 



Mix the crude carbon tetrachlorid with strong sulphuric acid in the proportion of 

 of acid to e\ery :i,IMM) cc of the carbon tetrachlorid. Shake this mixture thor- 

 oughly at freoiient inter\al- and allow to stand over niijht. Then run \\ater through 

 the mixture continuou-l\ . l.y means of a glass tube inserted to the bottom of the 

 bottle and .--.nn.'. t with the water tap, until thoroughly washed free from .i.-i.| and 

 impurities. Draw oft the water or upper layer by means of a siphon, the last por- 

 tion- li.-ini: removed as bu :'! by a pi|H-tte. Add an excess of soda solution 



and di.-til the carbon tetrarhlorid from it. 



II) "t tin- method is that a good blank can be obtained, the process of 



purification i- decidedly shorter, and it may be adapted to cruder carbon tetrachlorid 

 than can the present pn>\ isional method, thus allowing cheaper material to be used. 

 A blank should always be run on each set of determinations and if thin amount- in ihe 

 end to more than 0.2 to ().:; DC "t i. nth-normal alkali due to the carbon tetrachlorid, 

 the reagent i- not pure enough for this determination. The impurities present in 

 some samples of bichromate also gave trouble It i- absolutely necessary in this 

 method t hat reagents of all kinds -hall be entirely free from organic contami nut ion. 



In the extraction jxrtion of the method the followim: precautions are necessary : 



(1.) A shaking machine gives more regular condition- for extraction, each shaking 



ntinned for a period of two minu 



It i- of advantage to h.i\e perfectly saturated sodium chlorid and sodium -ul- 

 phaie, and for this purpo-c the solutions are kept standing over an excess of the salt 

 and continually agitated by a current of air. 



(3) It h. iM-rimeniully determined that a colder temperature gives a more 



efficient extraction. 



It i- al-o HIM es-ary to take special care to remove by complete washing with 

 sodium sulphate all the sodium chlorid from the carbon tetrachlorid extract ,,n account 

 of the formation of chlorin in the oxidi/ini; process with bichromate and sulphuric 

 a<i,| and the danger of this chlorin interfering with the titration by bleaching the in- 

 dicator The pre-ent pr- 1\ i-i -mil method calls for one washing with sodium sulphate 

 to accomplish this result, but it has been found by experiment that two are not suffi- 

 cient to remo\ e the sodium chlorid. < h\ the other hand, it may be that more than 

 two wa-hiiiL'- would abstract some of the higher alcohols, therefore it appears that the 

 it directions in the provisional method should be changed from one to two 



(5) It is necessary in carrying on the oxidation that the boiling of the carbon tetra- 

 chlorid with the oxidi/in^ solution should be slow and regular, and that a high con- 

 1 en -er should be n.-ed to insure the complete condensation of all the products. Espe- 

 cially i- rlii- slow boiling necessary in the following modified method which de- 

 pends on the estimation of the potassium bichromate reduced during the oxidation. 



T::I;T:I Hull. 1J O9 - 14 



