Mu 1c- >^< ,-, 1 W'^ c_ -i:ji S;: ^ ^ > „. V . - ■■:; , ■■''fr, *4 ', \ '%• ■ '*< ^ t- ’t* ' ,r . . / ■■ .. - ■ •; ,. , ‘r ,>C| ^,;i::ri?^v, : ;.ja. .. -'^ "' ’ ' i* ' ' : 'Vr- a^,.. .V ^v' A ' L\-, THE EFFECT OF 11 ME AND TEMPERA- TERE UPON THE BROMINE VALUE OF FISH OILS MY VIVIAN JONKS YAl.E THESIS FOK THE DEGREE OF BACHELOR OF SCIENCE IN LIBERAL ARTS AND SCIENCES COLLEGE OF LIBERAL ARTS AND SCIENCES UNIVERSITY OF ILLINOIS 1922 -.V . eiOV1UJt*qO 7Tl*«SV»!’ ••. 'X*v$ ^ :/ /'W ..^ •'jT i?;i . r4:tXX». «' TABLK OF COTv^TF^TTB. A 3 kn o wl e dgrn en t p ag e Introduction 1 Historic 3,1 u Theoretical 4 Experimental 9 Results — Table for Cod Oil 16 Table for Herring Oil 16 Table for Salmon Oil 17 Ta'^le for Sardine Oil 17 Table for Japanese Herring Oil 18 Table for Menhaden Oil 18 Conclusions 25 Bibliography 26 Digitized by the Internet Archive in 2015 https://archive.org/details/effectoftimetempOOyale ACKNOWLEDGMENT. I wish to express my sincere appreciation of the valuable assistance vihich has been received during this investigation, from Dr. George D, Beal, under whose direction it was carried out. INTRODUCTION. Natural oils and fats are widely distributed throughout the vegetable and animal kingdon, from the lowest organisms up td the most highly organized forms of vegetable and animal life, and are found in almost all tissues and organs. In plants they are chiefly found in the seeds and are associated therein with the starch to nouri^ the embryo. In the animal organisms, the oils and fats are mainly enclosed in the cel Lular tissues of the intestines and in the tissues nearest the outer skin. Considered chemically, fats and fatty oils consist chiefly of glycerol tri-esters (glycerides) of saturated fatty acids a.nd of related, unsaturated acids. The properties of the fats run parallel with those of the fatty acids and since this paper will deal with the unsat- urated fatty acids a few of their properties will be considered kiere. These acids contain one or more pairs of carbon atoms united by a double bond and are therefore, capable of taking up, oy direct addition atoms of halogen or halogen acid to form saturated compounds; the quality of the halogen absorption being dependent entirely upon these double unions. HISTORICAL. In about the year 1857 Cailletet' communicated to the Societe Industrielle de Mulhouse a titration process for the determination of the bromine absorption value of an unsa tur- ated f atty acid, to be thus performed: "To any given oil add a five per cent aqueous solution of caustic potash and agitate; then pour in an excess of a 33 1-3 per cent alcoholic solution of bromine; lastly, add a two per cent solution of turpentine until the color of the free bromine is completely discharged. The turpentine solution is supposed to be known in terms of the bromine solution, and thus the bromine absorption of any oil can be determined. If stand ard pure oils be at hand, their admixture can be quantitatively ascertained by this means." Cailletets' process had some very obvious disadvantage according to E, J, Mills^, the solution of bromine can not be preserved unchanged, probably even for a few hours; and the presence of water and alcohol must inevitably tend to promote oxidation of the oil under treatment. Again, Cailletet’ admits that the absorption increases in total amount with so moderate a rise of temperature as from 10° to ZOP c, and that a consid- erable amount of bromination occurs. It occurred, accordingly, to Mills^ (1379-1880), more particularly in connection with Hydrocarbon oils, that weak aqueous bromine alone might give -3- gentler and steadier results. This plan v/as accordingly tried with various samples of petroleum, the sample being agitated with 100 parts of water, and a one-tenth per cent solution of bromine added, with constant agitation, until potassium iodide and starch, used as an external indicator, gave a "^lue color. Under these circumstances constant bromine absorptions were obtained accompanied always with some bromine absorption. The chief difficulty in applying this otherwise satisfactory method consisted in the extreme amount of technique required to obtain uniform measurements. (1831) subsequently st^'gested agitating 0 s with excess of aqueous sodium hypobromite and hydrochloric aoid, and titrating the ultimate free bromine. This amounts to treatment with nascent as well as free bromine in the presence of water; so that the results are of a very composite character. As it was evidently desirable to confine the absorption to the formation of additive and not substitution products, it was decided to employ CS3 as the common solvent of the bromine and the substance to be titrated. Water was thus practically excluded and the reaction was marked by great regularity and considerable sharpness. Mills later substituted CCI4 explaining the change by saying that at ordinary temperature the solution of ©romine is more stable in CCl^ than in CS3. Thus it is to him that we owe the application of the bromine absorption value to the analysis of fats. His method is as follows: r-'T ; I . I' * ■ ■ ■: .-s.*?.sv . : '* y ^' . ' tv- ' •'^ m f >■ '.S,' ■ f\,' •;.'*• 6. -it;; c'j-; ^ Vr ' f f- V 't' ,^. - . -r:r ■= -,r , . ,■ \- Ij.' . ' ^ ;»v ■ m.. >'^v- r ,• I" X.v' !\ - ■ ** t' ‘i ' j * " ftii * ,,',v't:i .'# 1 * i>- "^ t ' ._. l ■• •• 4 . s’ a •* • * .. . ■'. ; ),,<*• ’ •■” !-j- * « • i"' I • ■ • ■« A 1 . .- -, , ‘ J,i . >' * V •‘■* r V. ■ V ■ V ■ % ir :r:. £ .no ;! yrv. w- -4 - ■■ ^v . :‘r'- % ^ ijs V-^ ■ •' t '■- ' ■ " ' : 'i ll d \t •: ..iv >- - i . j- '*Srft • '■’ '. - j A xi> • .; ■{. ,'••■ iii' ■? ir. ■ yi^> .. j ■ t\ <• ‘f-'v* i, '. .^1 I- (f I . • f Zi. zr. ‘.Z'Xijr^ V . S- .',' ■ -> '■• -■ r ' *:■ ''-A ;*!. i J ■ • . . .<■9 *.*• * 11 -* ■ - .iX . . . . v/.w V '.’• ** ^ ’ • •I " . ;i . ';.. V ' sji-X6..v’ if. y%-- , • - .... -:'V-? .• ♦? iJ'v. ^ I ' ' • > - ‘■''j '■ 'i'?’ ■ !l.. it: > ■ • c.' ” • I •" . . '. ■ ■ ' , ‘ ■ ■>.' -■ '■ .H ■ ‘ - .•■, ■■■■ ./r- : 3 - ;;-;■ • .*■ ?* if ■ ■ ;.; ^ .t’-"'.'vy, • * ' \ ^ "'''' ' c» !• i‘ ' i.X lx- f. ■ -■.'< ■'» ’♦k X,* t] t.V,- . Z..J oj!.^' -V t- ' f ' V ' • .. * ‘‘y..* ^ ^ V'f ^ .*.' rt\'1 vf ' ■’ - '* ' If- 'i IN if ; K. . u ' 1 .; ( .'.'• ’ t iff' . ,r A' . i A. 1. ^' V -4- .1 gram of oil or fat is tried tnoroughly and filtered, then dissolved in 50 cc of CCla and placed in a narrow-mouthed stoppered hottle of 100 cc . capacity. To this solution is added a standard carbon tetrachloride solution (about .006- .008 grams per c .c . ) until after a lapse of fifteen minutes, a coloration persists. The excess of bromine can be measured either by this color compared to that produced by a blank experiment or, more accurately by titrating back with a standard solution of B-napthol in carbon tetrachloride, when monobrom naphthol is formed. The bromine is calculated to -t,bsorb 100 grams of fat . The average probable error is said to be about .46 per cent . Mills emphatically stressed the fact that moisture must be rigidly excluded, since in its presence the bromine absorption number was found to be too high. Instead of B-naphthol, potassium iodide may be added, and the liberated iodine titrated with standard sodium thio-sulf ate . THEORETIC L . The bromine (or iodine) absorption value is defined as the per cent of bromine (or iodine) absorbed by the sample . This value is a quantitative measure of the proportion of unsaturated fatty acids, which, both in their free state and in combination with glycerol, have the property of assimilating halogens to form saturated compounds. # 5 * Two factors influence the result; (l)' the nature of the unsaturated acid yj hich is present, those with two doubl e bonds absorbing more than those with one even though the number of carbon atoms is the same, and (3) the chemical composition of the glycerides, those with the lower molecular weights absorbing a greater percentage than those with higher, where the degree of saturation is the same. It is hardly possible to allow dry bromine to act on an oil since it is absorbed with a more or less violent reaction and hydrobromic acid is evolved, hence the reaction m.ust be moderated by previously dissolving both the bromine and the oil in a suitable solvent . Carbon tetrachloride, as was first used by Mills® is generally employed as the solvent. In the course of the reaction, even when carbon tetrachloride is used, a small q uantity of hydro- bromic is formed, due to the substitution of hydrogen by bromine in the molecule of the fatty substance. Therefore the total absorption of bromine is due to both substitution and addition, and since greater amounts of substitution products are formed wjJh more concentrated brom.ine solutions, moderately dilute solutions should be used. The amount of substitution which has taken place can be determined by a measure of the hydrobromic acid formed and the true bromine a bsorption value, will be the total absorption - 6 - value minus the bfomine reacting by substitution. Mcllhiney'^ in 1894 determined the bromine addition and the substitution values as follows: ^Dissolve from .35 to 1 gram of the oil or fat in 10 cc of carbon tetrachloride in a 500cc stoppered bottle and add an excess of one-third normal solution of bromine. Place in ice after a few minutes in order to create a partial vacuum. Next slip a piece of rubber tubing over the neck to form a well and through this suck 35 cc water into the bottle. The contents are well shaken and about 1 0 cc of 30 percent potassium iodide e.nd about 75 cc of v/ater added. The iodine liberated is measured by titrating it with a standard solution of sodium thiosulfate and then calculated to bromine. The total amount of bromine added is ascertained by a blank test, and the difference betw'een the two equals the to tal bromine absorption. This is calculated to units in per cent of the sample taken. The contents of the bottle are now transferred to a sep'aratory funnel, and the aqueous solution separated and filtered. If the filtrate is blue, it is decolorize! by sodium thiosulphate and the free acid is determined as hydrobro- mic acid by titration w ith decinormal alkali, methyl orange being the indicator. The bro mine calculated from the hydrobromic acid and expressed in per cent of the sample gives the bromine substitution value, Tw ice this number subtracted from the total bromine absorption value furnishes the bromine addition number. In order to have a standard by which to measure the experimental portion of this paper, the iodine values of the oils I V ■ >* ' ^ 'Ti I I i ■ * ' ■^' M 4 -7- under exa mination were first determined. There are three methocb at present which are most commonly used for the determination of the iodine absorption value, those of H-Cibl, Wijs and Hanus. The first (Httbl)^ consists in the main in s.dding 50 cc of the iodine solution (consisting of a mixture of 26 grams of pure iodine in 500 cc of 95 per cent alcohol and 30 grams of mercuric chloride in 500 cc of alcohol, mixed in equal volumes and let stand 12 hours) t6 the sample, which is dissolved in chloroform in a special Erlenmeyer flaslfc. The flask is stoppered tightly and the gutter filled with fifteen per cent potassium iodide, then allowed to stand three hours — a blank being run at the same time — then 20 cc more potassium iodide is added and the contents titrated v/ith thiosulfate. The absorbed iodine is cal- culs.ted in terms of percentage of the weight of the sample. This method has been proven to be one of the most valuable ones to be used in the technical analysis of oils and fats. Long before any detailed explanation had been found for the reactions involved, the numbers obtained afforded a valuable guidance in the examination of oils. Experiments by Wijg led to a satisfactory explanation of the reactions occurring in the Hftbl solution. The important change which takes place on mixing the solutions of mercuric chloride and iodine is thus represented: HgCl3 +■ 2 12 «= Hgip. 4 g ICl K I .Vi?' ft*r <•« 4 ■.♦ • ‘ / , ■ ♦ 4. ^ ' H V. / V y/jJ \ i ■' j.i i.‘ ^ ,• I r' I . .• .• • ?»-' f ' .'(j .' ^ .t, It..: <. i r/- i ., X .1 A' ,nV '* V ’ • /-> ' f i . I 3 ^ - 8 - During standardization potassium iodide solution and water are added and the following reaction occurs, ICl ♦ KI * Kci 4 - Is Thus it is to he discerned that provided no other reaction has taken place, all of the iodine originally employed must he found in the hi ank test as iodine. Wijs' method differs from that of Kfiihls* only in that his iodine solution is prepared hy dissolving 13 grams of iodine in a liter of glacial acetic acid and then passing in chlorine gas until all is converted into iodine chlor- ide. This amounts practically to a tacit acceptance that iodo- chloride is the active agent. Hanui^ suggested that Wi^s’ method he modified hy using iodine bromide instead of iodine chloride. In this method the oil is dissolved in 10 cc of chloroform and 35 cc of the iodine solution is added. The mixture is allowed to stand for some time with occasiona,! shaking and then 1 0 cc of a 15fj solution of potassium iodide and 100 cc of water are added and the excess iodine titrated vfith standard thiosulphate until the blue color obtained with the aid of starch paste just fails to reappear . The results of this method agree very closely with those obtained hy Hiihls’ method but the results obtained by the Wijs method are higher than the others. EXPERIMENTAL . -9- EXPERIMENTAL . The experimental portion of this paper represents an attempt to determine the conditions necessary for the accurate determination of the bromine absorption values of fish oils. Tbe oils used for the work of this paper were Cod Oil, Salmon Oil, Herring Oil, Sardine Oil, Japanese Herring Oil and Menhaden Oil. The solutions used in the determination of the bromine absorption values were prepared as follows; 1. Standard potassium permanganate, used for the standardization of the thiosulphated solution. The potassium permanganate was standardized against chemically pure sodium oxalate in the usual manner. I, II. III. Wt . of Sample .1843 .1467 .3031 CCS Mn04 3^6 23.6 33.5 NF. KMno4 .0939 .0938 .0933 Average Normality Factor .0930 3. Sodium P^*^^sulphate, standardized against the potassium permanganate as follows: Approximately 10 cc of 15 per cent potassium iodicfe was placed in a glass-stoppered flask, then 10 cc potassium permanganate solution vir ere added from a burette, followed by ten cc dilute hydrochloric acid; and the solution allowed to stand ten minutes . - 10 - One hundred cc E^O were then added and the solution was titrated with the sodium thiosulphate-starch being used as the indicator. The results were as follows for the three standard solutions: I . II . (a) No. CCS KI«'Ino 4 10. 10. " " Na2Ss03 9.92 9.9 N. F. Na^SgOs .0937 ■ .0939 Average Normality Factor .0938 I. II. (b) No. CCS KMr .04 10.25 10. " " NagSoOs 10.05 9.8 N. F. Na3S30;3 .0948 .0948 Average Normality Factor .0948 III, 10 . 9.92 .0937 (c) I. II. No. CCS KMno 4 10. 10. " " NagSgOs 9.4 9.4 N. F. NagSgOs .0989 .0989 Average Normality Factor .0989 3, Bromine solution — prepared by diluting 5 1-3 cc or approximately 16 grams of bromine with enough carbon tetrachloride to make 1 liter of solution. 4. Potassium Iodide — 150 gras, to the liter. 5. Starch solution prepared by boiling 0.5 grams of soluble starch with 100 ccs. water. b -11- The following additional solution was used for the determination of the iodine values: 6. Hanus solution — 13.3 grms . iodine were dissolved in 1000 cc. of glacial acetic acid and enough bromine — about 3 cc — added to double the halogen content. In order to establish a standard for the determined bromine absorption values — and hence iodine values calculated from them — the iodine values of the six oils were first determined, according to the Hanus method. (see following page) Menhaden .3146 25 28.1 75.4 .0938 178.95 Average 178.79 ►TT« Ch Ch -12- w C/3 CO C/3 ffi o o p p p P P P CD CD o 8 83^ WOD 8 8 H M 8 8 a S' CD P CD P Oi Di 3 3 8 8 P 8 8 8 H« H* o o H. H* Oi *-i CD 8 CD 8 8 8 8 8 8 CD 8 H* CO 8 CD cq H* CO 8 CD oq CD CD cn cq Ol to • to to • H* • h-* • H* to to to <01 CJi <} cn oo CO cn GO CJi cn 01 <35 O 01 <35 <35 00 CO 00 CO M to to to to to to to to to to to Ol cn cn cn cn cn cn cn cn cn cn cn cn 45. 4^ 45. cn to cn CD M CD cn • • • • « • • e • • -d to to <35 CX) 00 00 o to to cn cn cn to to <3 -d ■<} <35 <35 <35 <35 cn cn cn cn cn CD CD CD CD cn 4^ 45. 4^ OO M CD CD CD CD 89 ‘ « o • O • o • o • O o • o • o • o • o CD CD CD CD CD CD CD CD CD CD 01 w Ol Cn cn cn K. ^ 01 Ol ^ cn 00 00 oo > 00 00 > 00 OO OO 00 > oo <5 < . A"-' I -13- The bromine absorption va,lues of the oils were obtained as follows: From .3 to ,3 grams of oil were weighed into a flask with a well fitting glass stopper, 10 cos carbon tetrachloride were then added in order to dissolve the oil and then 35 ccs of the bromine solution, and the flask packed in ice and salt for varying lengths of time as shovm belov/. A blank v/as run for every group of samples, being treated exactly the same as the samples. At the end of the specified tim.e, the flasks were removed from the container, 15 ccs of a fifteen per cent solution of potassium iodide and 100 ccs of water were added and the m.ixture titrated at once v/ith sodium thiosulphate, starch being used as an indicator. With the exception of the Menhaden oil, very little trouble was encountered in securing good results. The am.ateur is apt to shake the mixture too vigorously — and perhaps he may wait just too long before adding the indicator — but after a certain technique is secured the titration becomes quite accurate and the endpoint is easily discerned. After the liberated iodine had been titrsvted with tenth normal sodium hydroxide, using phenol thalein as the indicator. Practically no variation occurred in the amount of alkali used so it was shown that very little substitution was taking place and hence it was deemed unnecessary for the most accurate to calculate this . \ i fc . . . • • . • ••> if ,i .1 i n - 14 - In order to study the length of time necessary for the most accurate determination of the bromine absorption value, as compared with the iodine value, determinations were made in the following series one hour, two hours, four hours, eight hours, and sixteen hours. The temperature was that of an ice and salt mixture and varied from -10^ C to -16*^ C-being kept constant in each case by mieans of a heavily insulated box in which the samples were surrounded by the freezing mixture. The results obtained were calculated to the percentage of bromine absorbed, or the number of grams of bromine v/hich lOG grams of fat v\Tould absorb; then from this the Iodine number was calculated and compared with the theoretical value for the Iodine number. The following explains the determination and calculations involved . Cod Oil. First V/t . Second Wt . Sample Wt . Ccs Bromine solution Ccs llagSsOs Ccs Na 2 Sg 02 blank N. F. NagSgO^ T ime Br . No. 47.4377 47 .2077 .2300 25 . 35 . 61.3 .0948 8 hours 86.52 Calc. r. No. I Br. 137.35 -15- ' Now 61.3 - 35 = 26.3 vvhich represents the number Cos of .0948 N. bromine absorbed by .S3 grams of cod oil. One cc of normal bromine contains .07992 grams of Brs therefore the following gives the bromine value 36.5 X .0948 X .07992 s 86.53 the bromine absorption .23 value of cod oil standing in contact with bromine for 8 hours at the tem.perature of ice and salt . Since the previously determined iodine value is the measure of the fo maximum possible absorption, the iodine value is calculated in terms of the bromine value. Results : The following present the results by varying the time and the appended curves represent the same in, graphical form. •i ; ■',5^ i '■ N- •. ■ f f) .. .^■; j* t , V ^ ' A -> ^ V? tl -16- A' CD CD MWi-’rotoroMrorow p HMCOCDCDCOCOCOMCO P C/i 3 -OCJiCDCni-'cDOOcDCO 3 OCOOOCOihF^l-'‘^ H (0 CD 13 13 H !-■ H* j_j j_j H* o^a500094i>‘rfi^wroHt-' 3 c^cDOJOOrf^i^^cocoroi-' 3 CD CD l3'!3't5‘!3't3'|3'D'l3'tr!3' p'ortrp'p'p'p'trp'p* h3 CD 3 1 1 1 1 1 1 1 1 1 1 'X3 CD h-’MMHMl-'MHh-'l-' « 1 1 1 1 1 1 1 I 1 9 3 cocowroooMtocoro o tOIOCOCo^^;^^^to^ococo • o 03 a o Mwwrocowwroroto tj) cocococococororotoro CO oiaiaicjiuicnaiOTcnai 4 CJicnCJiCJiOiCnoiCJiCDCn GJ W o CD • o o o o CO •-S o Pi h-'MCOrOMIOINDl-’tOCO H* COCOCDCOMl-'CDCDCDCD 03 cDO^^^i‘C^Da^OJ-'CX)l-'M P ^^CJ3CDI-'CD^t^C0-'•••• P P oooooooooo p • P oooooooooo P • P tDcDCOCDCOCDCDCOCOCO CO 3 CD CO CDCOCDCDCDCDCDCD CO 3 (X)CDOOOOCDOOO^CO^P-^^l^ CD "t! cf CD o' tDcDcDCDyDcDcDcDCDCO . CD cDcDODCOOOCDCDOOOOCX) CO . p O O p CD CD M CD •3 -o 00 00-3<^COa)CDHCDh(s^ftSir0t033 3 0 0 tirt3't3'ffrrt3'D't3'!3'r3' 13 0 3 1 1 1 1 1 1 1 1 1 1 3 1 1 1 1 1 1 1 1 1 1 0 MMt-'l-'l-’H-'l-'l-'i-’l-’ 3 3 3 3 3 3 3 3 3 3 i_i(_ji_j|_itoroHooo • 3333tocooororo •'d o o rotococorororowrow 0 Cj aiaicjicnaicncricncncn cococotorotorocototo o cost ODODCncnCncnoiODCnOD 0 3 ^ 0 P • 8 DD •P 013* •d P o p « P Ui Qj o 0 H* MrOWWMWMl-' 0 3^^c^^^oolc^lto^oro^^ o 0 M--aOlOO!-'MOl01hf=> st:5 COOUlC-5 0305050 05 0! 0 0 p cDcOfJ^cDCnCOcDWMOJ ro O r03 otrooiti^cnODOD ta ro w cn C7) cn C/2H* Ol 03 OD 00 p 0 W 3 to •d O 1 w •:3 0^ M 3 o 05 (jQ o’ S o P o o' sd o O p ro 0 3 P O 3 cpcncncncnaicnaiuicji 1:5 Kn 33CD00CSDOD3333 p iV, 0 3 VO 8 CO 1 o prto OJ o 05 s s s: td oooooooooo p • OOOOOOOOOO p • COCDCDcDcDCDtDcDcDCD to M CDOcDcDtDCOCPcDCDcD to 3 CDCPOOCCi^^hP-COhf^ODCO C/D ^ COOOOOOOM^iSOOOOOOOO CO O CDCOCDOOOOOCDOOCOCD to . Pi CDCDCDCDOOOOOCDCOO to . Q. O H- o H* 05 p (35 d 0 0 M M H' M (-• M ^OOOOOCDCDCDCO DjP CDCDOOOOOOOOOOOOOOCO D3 d O cncjiai33 d . 3 • o' -OC^5C^3Cr30-0^3 0•^^;^C>5 o’ t^^M-^JCDOCnuiOOMOO Si0 305cnroooroto-ocy>3 0 0*3 o d 3 3 cn MHMMl-'l-’Hl-'Ml-' ooo 3333333333 o o <]OcnajC750>cjiaitf=>hi^ p. 05 0-5 05 OCI 0-1 05 to W p • 3-0 cnoCPOOCriOlcnOicDtD 3 3 OCD O ai tOc£)(7>03CnCOtDO5CD00 • 3aiC/iO<30D05CDt\3 3 • 00050000rOOCD '5s K ri rt> \ I‘f2 -? 1 - bSf y o 2 t 1 ''^ I '^5 4 o v» «u *: V. R) 5 oi ? ’s 6 <5 ■»~ •« 'S VJ <3 Vj V) i» til ■C: ^S v> u ::i 5 o < £t) r< cr V. Sdnit^/l dUJpof V c^ ir» ^ s 5 c^ > «> tv s 5^ /7f - 35- The preoeding ta’-les and curves for the various fish oils indicate that the bromine a";' sorption values of these oils increase with the length of time in which they are allowed to remain in contact with the bromine. In general the rise seems to take place rather uniforii.ly, a few ■‘^rea fis being noted — but since the ’whole increase is over such a small range — this might be due to slight differences in temperature, or a little loss in bromine because of its volatility. The oils with the lower iodine values seem to give the most consistent results, menhaden, as has been said before, giving the most trouble along this line, pro’^ably due to its greater degree of unsaturation. The increase in this case is also more marked with time. CONCLUSIONS . 1 . The Bromine absorption values of fish oils as determined vary with the time of contact with bromine. 2. The increase with time is related to the unsaturation of the fat, the less saturated sho'ving a more marked rise than the more saturated ones . 3. The conditions which come nearest the time of true absorption values as calculated from the iodine numbers are as represented with each oil — 16 hours at -10^ C . -r^ ^■’’•v V^ . *•*’ :;.V'A^Vj.*, V... • ' ,- •• . • * . ■ r7 ‘ ■ r-Jf r ■ ' ‘ V-' ■■• ■ - ■'- -V