Search

THE formation of more than one zone containing phosphate when the sodium orthophosphates are chromatographed on paper suggests that these spots contain different molecular species originating in proton transfer equilibria of the kind: where B−n is a base1,2. The speed of proton transfers, for example, the half-life for the reaction of hydronium and hydroxyl ions is 10−8 sec3, argues against this as such equilibria yield multiple spots only if the transition time is comparable with the time required for the formation of the chromatogram4. It is natural to question next if the zones are actually of different species since discontinuities in either the immobile or mobile phases can also produce the result4. Chromatograms of orthophosphoric acid and of the primary, secondary, and tertiary sodium salts were formed on 55 cm × 2.8 cm strips of Whatman No. 1 filter paper, acid washed, which had also been washed in 2 M acetic acid followed by three successive rinsings in distilled water5. The chromatograms were formed either with pyridine/ethyl acetate/water (the faster solvent) or butanol/water. Fig. 1 is a photograph of 10 µl. of 0.01 M trisodium phosphate (No. 321) and 10 µl. of 0.01 M sodium dihydrogen phosphate (No. 322) formed simultaneously with BuOH/15 per cent H2O. The origin was the vertical pencil line adjacent to the numbers. When revealed, the colour process was that employing ammonium molybdate and benzidine hydrochloride6. In certain select cases, the multiple spots were irradiated in the University's Triga reactor7,8. Both sodium and phosphorus undergo (n,γ)·reactions to form radioactive isotopes which decay as follows9,10

IN the procedure for the quantitative analysis of sugars, described by Flood, Hirst and Jones1, the paper strips containing the sugars are cut out and extracted by allowing condensed water to run down the paper. It has been found, however, that destruction of certain sugars occurs on boiling a solution under these conditions. To obviate this difficulty, we have devised a technique of extraction at room temperature similar to that used by Dent2 for amino-acids.

High-Voltage Paper Electrophoresis of Non-Volatile Organic Acids and their Mixtures with Amino-Acids

THE interesting suggestion that the oligosaccharides formed from sucrose may be the precursors of polysaccharides such as inulin and levan requires verification by the establishment of the identity and type of linkage of each of these compounds1. The recent discovery of their presence in plant extracts2 and human urine3 is a certain indication of their significance in metabolic processes.

High-Voltage Paper Electrophoresis of Inorganic Cations: Alkali, Alkaline-Earth and Other Metals

WE have shown1 that primycin, an agent of strong antituberculotic effect belonging to the group of guanidino antibiotics, is not identical with any of the known antibiotics. On the basis of its colour reactions and of its infra-red and ultra-violet spectra2, the molecule of primycin (C19H37O7N)n contains an N—H group and several O—H groups; it is inclined to molecular associations and it discloses adsorptive properties. On subjecting primycin to paper chromatography, the formation of a second spot was observed. This was due to the so-called ‘secondary adsorption’ of an accompanying substance present in minute amounts.

THE filter paper chromatographic procedure developed by Consden, Gordon and Martin1 has proved to be very useful in the qualitative and semi-quantitative analysis of amino-acid mixtures, and appears to have possibilities for extension to a variety of other compounds2, 3.

ELECTROPHORESIS has been of value in determining the physico-chemical constitution of cell-free extracts of micro-organisms1,2. These studies have been confined mainly to the bacteria and no such investigation of trypanosomal extracts has been made. Moving-boundary electrophoresis has been the method most frequently employed to analyse microbial extracts although the simplicity of paper electrophoresis would be of obvious advantage. The purpose of this communication is to describe the technique for paper electrophoresis and the resultant electro-phoretic patterns of trypanosomal extracts.

AGAR diffusion1 has been used for detecting antigens after paper electrophoresis. I have found that a precipitin reaction can be carried out directly on the paper.

THE detection and identification of pterins in small quantities was carried out by Schopf and Becker1 in the course of their classical work on this class of pigments. They employed a chromatographic technique using 0·004 N aqueous hydrochloric acid or 0·01 N methyl alcoholic hydrochloric acid solutions on micro-absorption columns of alumina or frankonite. The technique of paper chromatography has now been applied successfully to certain of the common pterins, and as a result a method has been elaborated for the rapid identification of very small quantities of these compounds such as may be obtained by extraction of individual butterfly wings. Indications that this method might be applied in the pterin series had been given by Crammer2, who mentioned that xanthopterin and leucopterin could be concentrated by a paper chromatographic technique.

MANNOSIDOSTREPTOMYCIN in some streptomycin preparations is conveniently detected by the paper chromatographic procedure of Winston and Eigen1. In the presence of high concentrations of salts such as in ion-exchange column eluates, or in fermentation broths, this procedure is not applicable: attempts to desalt such solutions, by mixed-bed ion-exchange, electrolytic de-ionization or solvent extraction, have proved unsuccessful. We have found it possible to separate streptomycin, mannosidostreptomycin and allied substances by paper electrophoresis, having developed (independently) a technique similar to that used for new antibiotic broths by Hosoya et al. 2, and King and Doery3.

AN automatic scanning technique employing a Geiger–Muller tube for the location and estimation of radioactive substances separated on unidimensional paper chromatograms has been described1. The application of this technique to two-dimensional chromatograms is tedious because a large sheet of paper has first to be cut up into the equivalent number of unidimensional strips and each strip scanned separately. To obviate this difficulty, an alternative technique to the conventional form of two-dimensional paper chromatography has been successfully developed for the application of two or more solvent systems to the same labelled mixture.

BY the use of rectangular strips of filter paper, rather than filter paper disks, in combination with a horizontal solvent flow at an elevated temperature, amino-acids are rapidly separated as round compact spots rather than circular bands1. This is desirable since such spots in many cases lend themselves to quantitative study by the maximum colour-density technique. Recently, such a procedure has been applied to the analysis of tryptophan in an alkaline hydrolysate of α-casein2.

IN a previous communication1 we reported that, using a butanol–ethanol–water–ammonia solvent, paper chromatography of diabetic urines and, in certain circumstances, of glucose samples, gave the normal glucose spot together with two extra spots of lower mobility. Subsequent studies of the inter-convertibility of the spots at various pH values have revealed that three spots, possessing the same RF values and chemical properties as those previously encountered, can be obtained when aqueous solutions containing glucose and ammonium acetate are chromatographed; the substances responsible for the two extra spots are formed when, following common practice, the solution under test is being dried on the paper in a current of warm air. The replacement of ammonium acetate by ammonium salts of other carboxylic acids and by ammonium carbonate led to similar results; in the case of the latter salt it was possible to observe that the compound responsible for the ‘middle-spot’ was formed prior to that responsible for the ‘back-spot’. From a larger-scale experiment in which glucose was heated with aqueous ammonium acetate, a product was isolated which was shown chromatographically to give mainly the ‘back-spot’. This product had a very small optical rotation, and gave the known crystalline α-diglucosylamine octa-acetate2 on acetylation; chromatography of the saponified acetate yielded one spot which coincided exactly with the ‘back-spot’. It seemed, therefore, most likely that this product was the equilibrium mixture of the α- and β-diglucosylamines prepared by Brigl and Keppler2, a view which was endorsed when these two diglucosylamines were used as reference compounds on chromatograms. We have shown further that glucosylamine, prepared by the method of Ling and Nanji3, is identical chromatographically with the ‘middle-spot’.

SEVERAL spray reagents are available for the specific detection of ketoses and ketose-containing sugars on paper chromatograms. Probably the most widely used are resorcinol and naphthoresorcinol, originally prepared as sprays in 2 N hydrochloric acid by Forsyth1. Afterwards, Partridge2 replaced the hydrochloric acid with trichloracetic acid in order to minimize attack on the filter paper. Heyrovsky3 has recently reported that β-indolyl acetic acid in trichloracetic acid is also a very sensitive ketose-specific reagent. In work in this laboratory on two fructose-containing disaccharides it has been found that all three of the above reagents, when prepared in trichloracetic acid, failed completely to detect these sugars on paper chromatograms. When the trichloracetic acid was replaced with 2 N hydrochloric acid strong positive tests were obtained. The results are listed in Table 1.

IN recent studies we were concerned with the detection of traces of nor-adrenaline in preparations of adrenaline derived from natural sources. Colorimetric methods1,2 proved to be of limited value with low concentrations of nor-adrenaline; but excellent results were obtained with paper chromatography, using modifications of the methods described by James and others3,4. The adrenaline samples are dissolved in methanol containing a minimum quantity of hydrochloric acid, and then transferred to paper strips for descending chromatography, or to circular disks of filter paper for chromatography by Rutter's technique5. Water-saturated phenol is used as the developing solvent in an atmosphere of hydrochloric acid4. The paper chromatograms are then dried and sprayed with an appropriate colour reagent.

WE wish to describe briefly a convenient paper chromatographic separation and identification of pyridine derivatives in use in this laboratory for some time which employs a modification of the wellknown Konig colour-reaction1. Leifer et al.2. have recently described the paper chromatography of nicotinic acid and nicotinamide, but the more involved microbiological method was used for defining the spots. In this laboratory the solvent used for the development of the compounds listed in the accompanying table is n-butyl alcohol saturated with 1.5 N ammonia. After evaporation of the solvent at room temperature, the dry paper is placed for fifteen minutes in a closed glass cylinder containing a few crystals of cyanogen bromide. The vapour pressure of this substance is high enough to provide a sufficient concentration for the first phase of the Konig reaction. The paper is then sprayed with a 0.25 per cent solution of benzidine in 1 : 1 ethanolwater. Amounts down to 5γ can be detected easily. It should be pointed out that although nicotinic acid and iso-nicotinic acid cannot be distinguished on the basis of RF values, they can be distinguished by distinctive differences in colour. This technique should be generally applicable to all non-volatile pyridine derivatives giving a positive Konig colour-reaction.

BENZIDINE has been found to be a useful reagent for sugar detection on paper chromatograms1,3 and for quantitative determination in test-tubes4. Horrocks3 first found that benzidine–acetic acid detected reducing aldoses on paper chromatograms, but ketoses were detected only at high concentration. Bacon and Edelman1 added trichloracetic acid to the benzidine–acetic acid reagent to increase its acid strength and claimed that it detected less than one microgram of glucose in free or combined form. Its sensitivity to fructose was considerably less. An acid is required for the formation of furfuraldehyde from sugars. Furfuraldehyde, or a degradation product of it, then reacts with an aromatic amine or phenol (benzidine in this case) to form a coloured product. It is important that the acid should he strong enough to cause the reaction to proceed, but, in the case of paper chromatography, it should not cause partial destruction of the paper or produce a dark background colour. For this purpose, organic acids (acetic, phthalic, or trichloracetic acid) have been used. However, when mineral acids at low concentrations are used, an improved reagent is obtained without paper destruction or serious background interference.

P work from this laboratoryl-S has demonstrated the heterogeneity of crystalline trypsin by paper and starch column electrophoresis. This article presents an electrophoretic method of greater resolution. Using this technique it was possible to demonstrate in a sample of crystalline trypsin the occurrence of fractions showing different ratios of proteolytic to amidasic activity. In the experiments recorded here a buffer of pyridine acetate of pH 4·9 (0,48 : 0·56 : 98·96 v/v of acetic acid: pyridine: water) was used for electrophoresis in a Durrumtype apparatus. To a strip of Whatman 3 filter paper of 12 x 50 cm, 0·12 ml. of an 8 per cent trypsin solution in 0·001 N hydrochloric acid were applied. An electrical tension of 450 V was used and the experiment was allowed to run for 5 h in the cold room. In these conditions at the end of the experiment the temperature of the buffer was 9° C. The strip, while still wet, was cut in 1-cm wide transversal segments and transferred individually to test-tubes containing 3ml. of a 0·25 M pyridine formate buffer of pH 2·7. This extraction was allowed to continue for about 20 h in the cold room. The extract of each tube was then analysed for protein content, proteolytic activity against azo casein and amidasic activity towards cebenzoyl-L-arginine amide (BAA). The protein content was determined by the FolinLowry4 method as applied to the 'auto analyseI" system·. Proteolytic activity against azo casein was measured by the method of Charney and Tomarelli". The hydrolysis of BAA was followed colorimetrically in the auto analyser'. Chromatography of trypsin in carboxy methyl cellulose was carried out according to the method of Liener8• The trypsin samples used in this work were purchased from the 'Worthington Biochemical Corporation ('Cryst. Trypsin, lyophilized, lot TL 747-58'), Freehold, New Jersey, and from Novo Industri ('Cryst. Trypsin Novo, batch 114-3'), Copenhagen, Denmark. The results obtained by the fractionation of the Worthn~on trypsin are shown in Fig. 1. The presence of three

VARIOUS methods have been described for the chromatographic separation of pyrethrin-type compounds. Alumina, silica and kieselguhr have been used as adsorbents for separation of ‘pyrethrin I’ from ‘pyrethrin II’1–3. A separation of cinerin I from ‘pyrethrin I’, and ‘cinerin II’ from ‘pyrethrin II’, has almost certainly been achieved by Ward2, using displacement chromatography on alumina. The ‘pyrethrins’ have also been separated by paper chromatography using alumina-impregnated paper1 and ‘Vaseline’-impregnated paper4. Hitherto, untreated paper has only been used with success for the separation of the ‘pyrethrin’ 2,4-dinitrophenylhydrazones, using light petroleum (b.p. 80°–100° C.) as solvent1. The present communication records a simple chromatographic system for separating pyrethrin-type compounds on untreated paper.

THE application of the principle of electrophoretic separation on paper to steroids is limited by their relatively low polarity, so that they do not readily migrate under the influence of an electric current, and also by their sparing solubility in aqueous media. By converting the steroids into derivatives with polar groupings, some separation has been achieved (Voigt and Beckmann1, 1953, using the hemisuccinates; McKinley2, 1955, using the mono- and di-phenylhydrazone derivatives). Our early experiments, begun before the publication of these authors' results, were along similar lines, and with somewhat similar results. It did not seem likely, however, that electro-phoretic separation would be so convenient a method as chromatography, which has been adapted by Lembart and Schneider3 for the separation on paper of mixtures of pure steroid conjugates. The natural occurrence of more polar steroid derivatives, the sulphates and glucuronides in urine, suggested to us that these might undergo electrophoretic separation on paper. Although the individual steroids could not be resolved, the two groups of conjugates were sharply separated.

SINCE the work of Lederer1 and Arden et al.2, the separation of inorganic cations by paper chromatography has been developed rapidly. The alkali salts, however, do not appear to have been studied, because of the difficulties of detection of alkali cations on the filter paper. We have recently discovered a simple electrolytic method of detecting the alkali and alkaline earth cations, and have studied the paper chromatography of the alkali salts.

IN a recent publication1 from this laboratory, a technique for the study of the chromatographic behaviour of enzymes on paper was described. The ease with which enzymes can be readily identified on the paper chromatogram by the agar-plate method enabled us to study the chromatographic behaviour of some of the important enzymes. In the course of preliminary experiments with amylases, phos-phorylases and phosphatases from various sources, we found that some enzymes did not move at all while others moved considerable distances on the paper. This observation prompted us to attempt the separation of enzymes by paper chromatography.

THE fluorescence of amino-acids and peptides on dry filter paper1 is of great value for locating these substances after paper chromatography, and has been found useful in quantitative work2 since it avoids destructive treatment with reagents such as ninhydrin. It also enables substances to be eluted for further investigations from sheets of chromatograms where gross irregularity of solvent flow renders the use of marker strips3 inaccurate as guides for cutting. Unfortunately, marked variations in fluorescence are often observed when identical chromatograms are run on sheets of No. 4 Whatman paper from different packets of papers or even on adjacent sheets from the same packet4, and batches of papers are occasionally encountered on which no fluorescence at all can be observed at normal nitrogen-levels. As Phillips1, quoting De Ment5, considered fluorescence of amino-acids on paper to be due to excitation of the acids themselves by ultra-violet light, it seemed possible that some impurity present in ‘bad’ fluorescent papers might therefore be quenching fluorescence. Sheets from a particularly bad batch of No. 4 Whatman paper could not be improved, however, by washing with water, dilute sodium hydroxide, hydrochloric acid or with solutions of 8-hydroxyquinoline or dithizone.

THE surface of filter paper, in contact with water, has a slight negative charge1, and this may account for the non-adsorption of certain colloidal electrolytes such as dyes of the ‘acid’ type, when developed with water, as these also carry a net negative charge in aqueous solution. By the application of principles similar to those described by Weiss2, a ‘primary adsorbate’ may be applied to paper, in this case being so arranged that the paper now has a net positive charge in contact with water. A suitable compound for this purpose is cetyl trimethyl ammonium bromide, the paper being soaked in 0.03 per cent aqueous solution and dried before use.

THE identification of non-volatile organic acids on paper chromatograms is usually dependent on the use of a suitable pH indicator which is applied to the developed and dried paper by spraying1. This procedure is inconvenient and an even distribution of indicator on the paper is difficult to obtain.

IN the amino-acid analysis of biological fluids using paper chromatograms1, the ninhydrin reaction indicates by means of a purple coloration not only α-amino-acids but also many other types of amino-acids and amines2. Spots that do not correspond in their positions on the paper to known substances are often found in these fluids. The circumstances sometimes preclude the possibility of isolating and analysing such substances, so that we may be limited to provisional identifications based entirely on paper chromatographic procedures.

CATIONIC detergents, such as cetyltrimethylammonium bromide (‘Cetavlon’) and cetylpyridinium chloride, have been used extensively in recent years for the fractionation of polysaccharide mixtures1–4. The quaternary ammonium cations present in these detergents form salts with acidic polysaccharides which often have solubility properties quite different from those of the usual metallic salts; for example1, cetyltrimethylammonium chondroitin sulphate is soluble in aliphatic alcohols up to n-pentanol and is insoluble in water. Many purification procedures have utilized the water-insolubility of these quaternary ammonium polysaccharide salts, but surprisingly, no serious attempt seems to have been made to exploit their solubility in organic solvents. This communication describes some preliminary experiments directed towards developing convenient methods for the separation of acidic carbohydrates, making use of the solubility of their cetylpyridinium salts in organic solvents. Current interest in these laboratories in the purification and identification of sugar sulphates, suggested that it would be useful if this property could be utilized in a method for the paper chromatography of these substances. It has been found that when a small amount of detergent is incorporated into the mobile (organic) phase of certain solvent systems for paper chromatography, the R F values for sulphated derivatives are increased by a factor of about ten, and some useful separations can be achieved. Two such systems are

IN recent studies on the paper electrophoresis of carbohydrate compounds, we have found that replacing the cellulose filter paper by paper made from glass fibres1 greatly facilitates the identification of sugars and their derivatives, especially those which are difficult or impossible to detect on cellulose paper.

A COLOUR reaction, known as Scherer's test, is extensively used as a specific, qualitative test for inositol. It is based on the formation of the coloured alkaline earth metal salt (usually calcium salt) of rhodizonic acid on the oxidation of inositol with concentrated nitric acid1,2. While studying phos-pholipids containing inositol we needed to identify accurately very small amounts of inositol by paper chromatography, or by spot tests on filter paper. P. Fleury et al. 3 reported a method in which they used mercuric oxide as a catalyst of the oxidation, and barium acetate as for the alkaline earth metal. This method, however, is tedious and sometimes gives uncertain results; the smallest quantity detectable is 5–10 µgm. This may be ascribed to the formation of yellow mercuric salt at the first stage of the reaction. A more sensitive, reliable and easier method was, therefore, devised by modifying that of Fleury. This modified method can, in addition, be applied to the detection of inositol diphosphate located on paper chromatograms. The procedure is as follows.

IN this laboratory the technique of circular paper chromatography has been modified in that small rectangular strips of filter paper are used rather than filter paper disks. This has resulted in the separation of compounds as round compact spots rather than circular bands. In addition, a combination of a horizontal solvent flow with an increase in developing temperature has resulted in a rapid increase in the mobility of the developing solvent coupled with a desirable increase in resolution.

TWO of the sulphonphthalein dyes, bromophenol blue1 and bromocresol green2, have been used widely for the detection of protein after electrophoresis on filter paper. The uptake of bromocresol green by protein under these conditions has been studied recently by Franglen and Martin3. Extending this work, the purity of other sulphonphthalein dyes has been investigated with paper chromatography and filter paper electrophoresis; the methods used and the results obtained are described below.

ATTEMPTS at the separation of higher fatty acids by paper chromatography have not, in the main, been successful, although Isherwood and Hanes1 have studied the problem in relation to a number of straight-chain fatty acids, and Nunez and Spiten2 have evolved a technique, using paper impregnated with olive oil, for chromatographic separation of such compounds. Apart from other considerations, the low aqueous solubility and the low dissociation constants of these acids presented the problem of finding an indicator which would be sensitive enough to enable their presence to be detected on paper. It has been found that the methyl red/bromthymol blue indicator of Duncan and Porteous3 permits of the detection of small amounts (for example, about 80 µgm. of C16) of acids up to C18 on paper. Considerably more care is required, however, in the colour development with the higher acids than with the lower members of the series.

IN work concerned with the composition and estimation of neurohypophysial hormones in the posterior pituitary, the hypothalamus and body fluids, it would be desirable to have a chromatographic procedure which permits the identification in and recovery of small amounts of the active peptides from impure extracts. It would also be useful to know whether the substituted synthetic oxytocins which have recently been prepared1 can be distinguished by a simple chromatographic method. Taylor2 separated oxytocin, and arginine- and lysine- vasopressin by ion-exchange chromatography on resin columns. He used du Vigneaud's highly purified peptides, and the smallest amounts applied were equivalent to 0.6 mgm. of this material. Benfey3 purified Stehle and Fraser's4 preparations of ox posterior pituitary hormones (200 U./mgm. pressor activity and 250 U./mgm. oxytocic activity) by paper chromatography using ninhydrin as the colour reagent. The smallest amounts of the hormone preparations used were 1 mgm. However, since only terminal amino-groups react with ninhydrin, it is obviously disadvantageous to use this reagent when attempting to visualize very small quantities of the posterior pituitary peptides. Rydon and Smith5 have shown that micrograms of peptides of all kinds, including cyclopeptides, can be detected by chlorination followed by spraying with a starch–potassium iodide solution. Reindel and Hoppe6 improved this procedure in two ways. First, before exposure to chlorine, they moistened the paper with a mixture of watery ethanol and acetone, which prevents background staining. Secondly, they treated the N-chloropeptides with a mixture of potassium iodide and either benzidine or o-tolidine in acetic acid and thereby increased the sensitivity of the method.

THE possibility of combining the advantages of circular paper chromatography with those obtained by using an automatic fraction collector for separation and isolation of substances from a mixture has been investigated. The essential features of the technique consist in the application of the method of preparative circular paper chromatography1 for the separation of the substances on a paper pack, followed by the elution of the components by means of a device for facilitating the flow of the solvent from the edge of the paper pack. Fig. 1 is a diagrammatic sketch of the apparatus devised.

IT is usual to make solvent troughs from sealed lengths of glass tubing in which the opening for the papers is cut1, ground2,3, or pulled out of heat-softened glass4. These methods are difficult and usually involve many fractures. Troughs have also been made from stainless steel5,6and from pressed sheet metal coated with resistant enamel7, but these are not easy to make and are relatively costly, especially for workers in areas where financial grants are meagre and materials and apparatus difficult to obtain.

THE method used by Consden, Gordon and Martin1 for the qualitative analysis of amino-acid mixtures, and by Partridge2 for sugars, employing filter-paper partition chromatography, was applied to the determination of some keto-acids. The keto-acids were characterized as their 2,4-dinitrophenyl. hydrazone derivatives ; this gave the advantage of manipulating coloured compounds, which showed their position on the paper without further development.

IN previous publications1–4 from this laboratory, the use of the agar-plate method for the detection of enzymes has been described. This technique has now been adapted to the location of enzymes on paper. Preliminary investigations have shown that enzymes can be readily identified on the chromatogram by means of this technique.

LEIFER et al.1 have reported the paper chromatography of nicotinic acid with the use of microbiological assay. While our work was in progress, Huebner2 described a spot test for detection of some pyridine derivatives. The present work describes sensitive qualitative techniques developed in this Laboratory for the detection of tertiary and quaternary pyridinium compounds. The laborious procedure involved in the microbiological method of Leifer et al.1 is obviated by rapid colour or fluorescent tests. For the colour test, advantage was taken of Konig's3 reaction, namely, the formation of coloured compounds when tertiary pyridine derivatives are treated with cyanogen bromide and a primary aromatic amine. The fluorescence test allows the detection of quaternary nicotinamide derivatives. Use is made of the formation of bluish-white fluorescent substances when quaternary pyridine compounds with a side chain CO—NH—R in the β position are treated with methyl ethyl ketone and ammonia4.

OWING to the frequency with which caffeine is used for the illicit medication of racehorses, the positive identification of µg quantities of this compound is a matter of some importance. Although a caffeine spot on a paper chromatogram exhibits a deep blue fluorescence under ultra-violet light, and may be eluted for subsequent identification by spectrophotometry1 or by crystal tests2, in common with most other xanthine derivatives caffeine does not react with the usual spray reagents, and there is need of a colour reaction which can be carried out directly on the paper.

DR. M. D. GWYNNE has described an easily made solvent trough for use in descending paper chromatography1.

WHEN ethyl pyruvate and 4,5-diaminopyrimidine were refluxed in water, between pH 1 and 3, a solid of molecular weight about 320 was formed which was seen as two spots1 after dissolution in 0.1 N sodium hydroxide and chromatography on a sheet of Whatman No. 1 paper in water at pH 5.5 or 8.5. The spots (Fig. 1A) had RF values of 0.55 and 0.65 respectively and appeared identical when viewed in ultra-violet light. When eluted, and rechromatographed, each spot generated only one spot, which corresponded in RF to the spot from which it was eluted (Fig. 1B and C).

THE observation and analysis of isolated cell fragments gives much information on the way in which the more complex structures are put together. The usual method used for separating cell fragments is differential centrifugation, in which the difference in size and particle density is the determining factor. Albertsson1 has studied the separation of cells and cell fragments using column chromatography. This is a method of interest since it separates the cell fragments on the basis of the physical and chemical properties of the surface of the particles. This communication points out that the well-known technique of paper chromatography can be extended from the separation of molecules to the separation of cell fragments of very large molecular weight.

PHENOL and collidine have remained the most widely used pair of solvents for two-dimensional paper chromatography of amino-acids since their introduction by Consden, Gordon and Martin1. The use of the latter solvent, however, is attended by certain disadvantages such as double spots and haloes2, an offensive smell and possible toxic effects3,4 and, in some countries, difficulty of supply. Alternative solvents have been suggested, for example, butanol – acetic acid5, acetone – water2 and pyridine – amyl alcohol6. Some defects of the latter system have recently been reported7.