REPORT BY Ma. DAVID HOOPER, GOVERNMENT QUINOLOGIST, MADRAS, ON THE RESULTS OF ANALYSIS OF HEMP DRUGS.

I have the honor to forward the results of the analyses of samples of hemp drugs submitted for chemical examination by your Commission, together with those co.11ected by myself and others in this presidency.

This is the first time that a systematic examination of hemp drugs from different parts of India has been conducted. Previous investigators have contented themselves with sealching in commercial samples for the active principle, without first satisfying themselves as to the variation in the composition of the plants grown in different districts and under different circumstances.

The active principle of hemp is contained in, and intimately associated with, the resinous secretion which is found in the leaves of the plant, and which is formed in abundance in the flowering tops when the cultivated female plants are grown by themselves. The resins in this secretion are associated with one or more alkaloids existing in small quantities and not possessing the peculiar physiological action of the drug. The chief object of the analysis was not to determine the alkaloid, as in examining tobacco for nicotine, but to separate the resinous matter in as pure a condition as possible, taking the precaution to free the sample from excess of seeds yielding fixed oil.

Other objects sought for in the examination of Indian hemp were the föllowino.: • determinations of extractive's directly soluble in spirit and in water ; determination of nitrogen ; search for sugar, and estimation of that body when present ; a proximate analysis, showing the amounts dissolved by ether, and subsequently by rectified spirit and water, and the crude fibre and ash contents ; nature of the ash, solubility of ash in water with estimation of alkalinity, solubility in acid', amount of insoluble or sandy matter, analysis of ash of ganja and bhang ; proportion of seeds in genie and bhang; composition of the seeds; nature of the 'resinous extractive.

The literature on the subject of the chemistry of ganja has been referred to and digested, •and. Dr. Prain's " Report on the cultivation and use of ganja" has been read with interest ; but before offering any remarks I will first enumerate the samples of drugs received, describe the methods of their analysis and give the tables of results.

The samples of ganja.—In order to compare the composition of all the samples with those of acknowledged superiority, such as o- ganja, specimens were obtained through the Commissioner of Excise from Naogaon in ajshahi district. The four kinds—large flat twig, 'small twig round and chur were typical specimens and. were received in good condition. Surat, No. 1, was in matted heads with a fair amount of seeds. Surat ganja, No. 2, was in a loose dusty condition with many seeds and some stones. Besides these impurities, it gave a grass green coloured tincture, characteristic of bhang, and as leaves predominated it was placed among the bhang. The two kinds of Sind ganja came from Karachi ; they were labelled "first kind known as Panvel ganja," "second kind or Sholapur ganja." The first specimen had a good odour and was broken up into smaller pieces than the second, which was in clean matted heads. The Commissioner of Excise, North-Western Provinces, sent some pieces of Baluchar and Pathar ganja used in Allahabad, but not grown in the provinces. These fragmentary samples did. not admit of a full analysis being made. The North-Western Provinces supplied. two illicit aganjas—one from Ghazipur, and the other from Basti. The latter was a small sample " obtained by a fakir from the border of the district or from the adjacent Nipal terai." It was rather seedy and dirty in appearance. The Ghazipur specimen was of a good green colour, but the leafy bracts were thin, and the tops were not agglutinated as in other samples. The Khandesli ganja, bad the Dhulia post mark on the parcel, and was representative. The Satara drum was labelled " ganja flower tops"; it was seedy, it contained a few insects and had only a por odour. The Ahmednagar specimens were interesting in having a different commercial value attached to each. One was sold for Rs. 55, the second for Rs. 50, and the third for Rs. 48 for three Bengal monads. The retail price of each was 12 annas per pound. There was not much difference in their appearance, except that perhaps there was more stalk in the third kind. The Nasik ganja was in matted heads with fair odour, and almost entirely free from seeds. The Solahpur sample was clean and had few seeds; its composition was different to the Solahpur genie imported in Sind. The Bijapur ganja had a few seeds, a slight musty odour and a brownish-green colour. Of the two specimens from Niniar distriet, one was marked "cleaned ganja raised from local seed," and the other "cleaned ganja raised from Dhakalgaan seed." The Hyderabad drug was of a good colour and fresh. With reference to the Madras samples, the two kinds purchased in the Potacamund bazar were called " javaj " or " Javathu " and "country genie." The first is considered the best kind, and from its name very probably came from the Javadi hills; the second kind came from Coimbatore. An authentic sample of ganja from Coimbatore was obtained with the view of comparing it with the " country ganja." The Tanjore sample was sent by Dr. Mootoosawmy of that city, who stated that all the ganja was " imported from Chittore, Vellore and a mountainous village in the latter district called Kaleyimbody." Bangalore also appears to be supplied with ganja from Vellore. The sample from Madras city purchased by the Commission was said to have been imported from the Jawadis. It had a cummin-like odour. Mr. Benson, Deputy Director of Agriculture, obtained several specimens of Ganjam ganja which varied very much in colour, odour and proportion of seeds. My aealysis was made on a mixture of the better kinds containing few seeds. I am also indebted to Mr. Benson for procuring for me sample of the ganja prepared in Daggripad in the Kistna district. The small leaves alone with the flowering tops were very numerous, linear, with revolute margins, and very brittle. The fruit was set, and the bracts were covered with brownish glandular hairs, and the whole was very fragrant of southernwood. This sample is remarkable in yielding the largest amount of resin of all the others, including the Naogaon ganjas.

The samples of bhang.—With the exception of the samples from Bhagalpur and Mongbye which were sent by the Collectors of those stations, all the others were forwarded by the Commission. They varied in odour and in amount of seeds and impurities. Some had entire leaves in others they were broken up to a coarse powder. The bhangs from Bhao;a1pur and Monglyr had a marked peppermint odour, those from Bijapur and Sholapur bad only a slight fragrance. The Khandesh sample was very much broken up and contained some seeds, while that from A:nballa had its leaves twisted up like green-tea, and was remarkably free from seeds and foreign bodies. The samples from Surat and Hyderabad w ere much contaminated. In these cases the Surat bhang was powdered up and analysed just as it was received, and the Hyderabad mixture was sifted and the cleaned leaves only were used for analysis. The wild bhan„o• from Assam had broad thin leaves, and a mousey odour. The leaves from the cultivated Rajsbahi plants were lighter in colour, thicker, linear, with revolute margins, and a "herby " smell.

The samples of wild hemp plants were allied to those of bluing, but being in smaller ivantities they were analysed by a different process. From Basti, North-Western Provinces, three interesting specimens were submitted—a sample of bhang with starninate flowers, female hemp called the " ganja plant ", and male hemp called the " bhano"i plant." From Gonda two plants representative of the sexes were sent. One was said to be obtained from Ammar-pur, 21 miles east of Gonda, and was bearing staminate flowers ; the other was a female plant from the same village in which the young fruits were developing. The specimen of wild hemp plant from Jaunpur, North-Western Provinces, consisted of leaves without any flowers or seeds and was stated to be used as ganja for smoking; but nothing definitely was known about this. Bliang does not seem to be sold in South India as a commercial article, but the leaves of the wild plants grown near houses are used by the natives for smoking. A specimen of sucks plant was obtained from Pykara, on these hills, and its analysis resembles very strongly that of the wild bhang from Assam.

The samples of charas.—In Amballa district, cbaras " mashak " sells for Re. 1-8-0 per seer. It has a dark olive-green colour, tough consistence and a peculiar fragrant 'aroma. The Amritsar, Delhi and Bombay drugs are very much like the above, although called by different names, such as " mashak ' , " bhara" and "dust", and sold at prices ranging from 12 annas to Re. 1-9-0 per seer. Gwalior charas occurred in black balls from the size of a pea to that of a nutmeg. The sample from the wild plants in the Kumaon terai were spindle-shaped and about l inch long. They contained seeds and much vegetable matter, and were very probably obtained by rolling ganja in the hand. The Himalayan charas from Kumaon cultivated plants was in the form of large balls made by massing about a dozen small balls together. The Himalayan charas made from cultivated plants in Gurhwal was in two forms—flat, square shaped pieces about j inch across, and round disc-shaped flattened pieces about 1 inch in diameter, with a hole in the middle by which they were strung together on a string. These had a heavy tobacco-like odour, and were black in colour. The Yarkand charas was a small hard cake weighing a few grains, sent in the samples of ganja and bhang from Allahabad. The Almora charas was a portion of a hard ball in which seeds and other vegetable debris were present. The two specimens from Nipal had the odour of musk. " A " was a sampla of 1892 manufacture of good quality, and occurred in small rounded discs like Pontefract cakes. Sample "B" was two years old, and in cylindrical pieces, black and hard; ,it was called Shahjahani charas.

The methods of analpsis.—The samples of ganja were broken up by band, the stalks removed, and the loose seeds rejected, and the matted heads at the extremities of the smaller twigs were only taken. This " chur " was then powdered and made to pass through a sieve with 36 meshes to the linear inch, and the resulting powder was kept for use in a stoppered bottle. A tincture was made by macerating a weighed quantity of the powder for 98 hours with rectified spirit, percolating to exhaustion, and evaporating to dryness in a water-oven. The result would represent the amount of extract, the chief preparation of Indian hemp adopted by the pharmacopceias. The infusion was prepared by allowing five grammes of the powder to stand in warm water for a few hours. If cold water is used for the extraction, it requires more than 21. hours for complete saturation, and before that time the extract shows signs of fermentation. One portion of the watery extract was evaporated to complete dryness, another portion was precipitated with solution of acetate of lead to separate the organic acid, and the third portion, if sugar was present, was used to titrate a measured volume of Fehling's copper test. The nitrogen was determined by combustion with soda-lime, and existed in the plant mostly as albuminoids, although some was yielded by the alkaloids, and in some cases by a soluble salt of ammonium. The proximate analysis of ganja and bhang was commenced with ether, which is the best solvent for the active resins, and dissolves very little extraneous matter. The subsequent action of rectified spirit (sp : gr : o• 83) removed a resin acid, alkaloids, and in some cases a saccharine body reducing Fehliug's solution. By adding together the ether extract and the resin acid, the result equals the amount of washed resin obtained by the direct spirit extract. If the spirit extract in the proximate analysis is much over 2 per cent., it points to the presence of sugar ; absolute alcohol, on the other hand, does not dissolve this sugar so easily. Dr. Prain used petroleum ether for the first solvent to act upon the drug, and the dried extract obtained by this liquid was returned as "fixed oil, etc." If this were the case, then all the extracts of Indian hemp would consist mostly of fixed oil, whereas the extracts from all ganjas consist, as I shall show presently, of resins with a small proportion of fixed oil. Petroleum ether is a good solvent of the resins, and is of no use in effecting a separation between them and the oil. The best method is to separate the seeds from the sample before the analysis, and so exclude the source of the oil. A misconception seems to have arisen from a statement made by Ron; a French Chemist, in 1887, that the ethereal extract is inert. If the ethereal extraction follows that of petroleum ether' then, perhaps, the product is inert, but direct exhaustion of the drug with ether certainly does not yield inert resins. Dr. Praia estimated, by an indirect method, the essential oil in some samples of ganja, and obtained some high results, over 6 per cent, in most cases. Indirect methods of estimating volatile oils give higher figures than by direct distillation, and are not to be depended upon. I have recently distilled one of the most fragrant ganjas, that from the Sistna district, and it yielded less than one per cent. of colourless essential oil, having the peculiar odour of the drug. The essential oil in the other samples has not been determined. The remaining analytical methods do not require any further explanation.

Seeds in ganja.-Samples of ganja contain variable amounts of seeds, or, properly speaking, fruits. A fresh sample of " sadai " ganja from Coimbatore contained a large number of ripe fruits, and as it was said to afford commercial gan ja, I carefully dried some of it in the sun, and separated the seeds from the other vegetable structures. The seeds weighed 565 per cent, of the dried sample, but if such a sample had been manufactured by being rolled or flattened out before it was dry, many of the seeds would have been lost. Some of the commercial samples of bhang contained much seed, notwithstanding the opinion that bhang is the produce of the male plant. The specimen of Hyderabad bhang, when cleaned, yielded 44 per cent. of seeds, with some stones and other impurities. In tile case of a similarly impure specimen from Surat, the drug, after the larger stones had been removed, was powdered up with the seeds and other impurities. The analysis of such a sample shows an increase in the ether extract due to the oil, and an increase in the mineral matter due to the sand, and is far from uniform with the other specimens. Twelve commercial samples of Ganjam ganja, very different in appearance, were found to contain from 5.1 to 381 per cent. of seeds, or an average of 20 per cent. As the seeds contain 25 per cent, of fixed oil, five' parts of the ether extract obtained from an average sample of such ganja, without removing the seeds, would consist of oil. In preparing samples of ganja and bhang for analysis the seeds that fell out in breaking up the drug were rejected altogether, and I do not think that, in the foregoing tables of analysis, more than a very small proportion of the resinous extract consisted of oil from the seeds.

Composition of the seeds.-The cleaned seeds, separated from the Hyderabad bhang, were reduced to fine powder, and analysed with the following results :

The most recent and complete analysis of hemp seeds has been published by S. Frankfurt (Lamb° Fersuchs. Stat. 43.113-182 Journ. /,em. Soc., March 1894, p. 113). This examination, it must be pointed out, was made upon European seeds

The extracts of the seeds do not reduce Fehling's solution without previous boiling with an acid; the sugars found in the hemp drugs were therefore not derived from the seeds. The percentage of nitrogen (3'57) in the seeds is higher than that obtained in the ganjas and bhangs, a result which would be expected by any one having an acquaintance with vegetable physiology.

Sugar in hemp drugs.—The occurrence of sugar does not seem to have been noticed by other investigators. It was detected first in the samples sold in the local bazar at Ootacamund, and as it was absent, from some Bombay samples, and appearing in traces in Bengal ganjas, adulteration was suspected. But as sugar occurred in all the Madras samples, and in pretty .uniform quantity, and in the cultivated leaves from Rajshahi and the wild plants from Assam and the Nilgiris, it could not be considered anything else than a natural constituent. The nature of the sugar, besides being amorphous and allied to glucose in reducing potassio-cupric tartrate, was net investigated. An extract of ganja containing sugar, if prepared with rectified spirit, would after a time separate into two portions, a greenish-black resinous mass, and a brownish soft substance of a liquorice-like odour and sweetish taste. The presence of saccharine matter would not be desirable in a medicinal extract of Indian hemp, as it would dilute the active resinous matter; on the other hand, if present, it could easily be removed by washing with water.

Organic acids in hemp.—Thero is no tannic acid in any of the hemp drugs. The acid removed from the aqueous solutions by means of lead acetate resembled citric acid more than anything else, but as a loss was experienced in separating this in a state of purity, other acids might have been removed in the mother-liquors. Oxalate of calcium was present in all the samples and accounted for the large quantity of calcium carbonate present in the ashes. • The larger the amount of solid aqueous extract in the drug, the higher was the proportion of the organic acid; but this rule is interfered with when sugar occurs in the extract. The percentage of oxide of lead in the lead compound of the organic acid ranged from 55 to al per cent. The colouring matter of the extract was by removed Y lead acetate, but another substance giving a transient purplish colour with ferric chloride was left in solution.

Nitrogen.—The percentage of nitrogen ranged from 1.96 to 3.18 in the ganjas, and from 1.91 to 3.17 in the samples of bhang. In the case of ganjas the high percentage of nitrogen is generally accompanied with a large percentage of resins, but there are some remarkable exceptions to this rule. The best of the Madras samples, judging froRthe yield of resins, do not contain so much nitrogen as some Bombay samples inferior in resin contents. The Ahmednagar sample of the best quality and richest in resin contains the lowest amount of nitrogen and vice versa'. The two samples of Sind and those of Nimar have the higher proportion of nitrogen associated with the lower proportion of resin:. The Naogaon ganjas contained alkaloids and ammonia in the aqueous solution of their spirit extracts. Ammonia was not detected in many of the other samples, and was absent altogether where the nitrogen was low. Ganja requires high cultivation, and no doubt in Naogaon and Daggupad, where. the growth is well sup. erintended, manure is largely used. Both the ganjas and bhangs contain more nitrogen than the average green fodder' so extensively used. The crude fibre contained a considerable amount Of nitrogen existing. as insoluble albuininoid matter, as ammoniacal fumes were given oll each time it was ignited.

The resins.—The best gauge of the resinous contents of the samples is to observe the percentage of "ether extracts" in the foregoing tables ; the direct spirit extracts contain, as I have pointed out, variable amounts of water-soluble matters in addition to the resins. The Ristna ganja is the best in Madras, and the Madras city sample from Javadi hills is the worst.

The Naogaon ganjas, as a whole, are superior to the other supplies, with the exception of the Kistna sample, giving over 25 per cent. of resins. It is strange that in appearance as well as in composition the " first kind" of Sind •ganja imported from Panvel should be inferior to the "second kind" imported from Sholapur. Another aspect of the ganja trade is the difference in the composition of samples said to come from the same place. The Sholapur ganja sold in Sind contains 2111 per cent, of resins, while that from the district or town itself yields 14046 per cent. The ganjas from Ootacamund, Bangalore, Tanjore, and Madras city come from the Javadi hill, yet they yield 21.88, 18.5, 17.37, and 13.92 per cent. of resins respectively. The " country ganja" imported from Coimbatore and used on these hills contains 11.61 per cent. of realms, but a sample obtained from near the town of Coimbatore gave 1844 per cent. We may learn from the analyses of the samples from Nimar in the Central Provinces that, whether from local or imported seetl, the ganjas grown in the same district and under the same circumstances yield similar products. With regard to the nature of the resinous extract of ganja, it is insoluble in water, but soluble in alcohol with a neutral reaction. A small itiaount of resin acid, about half per cent., was present in all the samples of g-anja and bhano.. A certain amount of the resins dissolves in warm potash solution (one per cont.) ; a still further quantity is dissolved by treatment with alcoholic potash, evaporating to dryness, and washing the insoluble resin. Ganjas treated in this manner give from 50 to 60 per cant. of purified resin, and bhang-s from 40 to 50 per cent. It is thus shown that the resinous extract of the hhangs is not equal, weight for weight, to that of the ganjas, as it contains a smaller proportion of purified resin which has been proved to be active. The bbangs contain from 8 31 to 12.63 per cht. of resins, or an average of about 10 per cent. which is one-half the amount yielded by average samples of ganja. Charas, it will be seen, contains, on an average of fifteen samples, about 40 per cent. resinous extract, which is doubfe the quantity afforded by ganja.

The ash of gaoja and bhang.—The amount and composition of the ash of hemp drugs is more uniform than those of tobacco. The proportion of total ash of ganja is generally undei- 20 per cent. If it is over this proportion, a glance at the figures under "insoluble ash " will show that it is due to an excess of sandy residue. The soluble ash (alkaline salts) does not exceed 3.43, nor fall below 1 67 per cent. The ash soluble in dilute•hydrochloric acid (lime salts, etc.) falls between 1357 and 8.68 per cent. The sand (insoluble ash) is the most yariable constituent, as it ranges from 8.53 per cent. in the Tanjore sample to 2.18 per cent. in the Kistna drug. The alkalinity of the ash slim-Vs much attachment for the figure 1. The proportion of total ash in the samples of bhang in every case exceeds 20 per cent., a result which is due to the Ihrger amounts of lime salts, as 'well as sand, than are found in the specimen of ganja. In the analyses of wild hemp plants, the male plants appear to yield more ash than the female plants, but the latter more resin. The composition of the Basti " bhangi plant " is not in 1:ee1ing with the above statement, and is abnormal in yielding so much extract to spirit and so small a quantity of ash.

There is a great similarity in the composition of the ash of ganja and bhang. They are both fairly rich in calcium phosphate, and the ash from ganja contains more potash than that of the bilking.

Remarks on charas.—Charas has very seldom been examined chemically-. Flückiger and Hanbnry, in " Pharmacographia," second edition, page 550, report that charas yields from one-quarter to one-third:of its weight of an amorphous resin, and Dr. Pram in his report refers to samples yielding 75 and 78 per cent. of resin. It will be seen from the results of my analyses of North Indian Samples that the average yield of resin is 40 per cent. The highest is 46.5 per cent. in "mashak " charas from Amritsar; and the lowest is 22.S per cent, in a sample of charas made from wild plant grown in Kumaon. The latter sample has the composition of a good sample of ganja. The samples from Amballa, Amritsar, Delhi and Bombay are remark-' ably uniform in their comppsition and physical characters. They had a similar odour, consistence and colour, and all contained a large amount of sand. The Gwalior and Himalayan sarx,ples had portions of leaves and sometimes, seeds mixed with them, and left a quantity of vegetable residue after extraction with spirit, but there was not so much sand present. The Yarkand charas bad some 'carbonate of calcium present, and the ash soluble in hydrochloric acid was consequently high. There was very little in the resin contents and appearance of the Delhi samples to determine their. money value. The bast quality had less resin and more sand than the other two kinds ; it was more plastic and contained more water than they did. If. alkaloids were the active principle of charas, then one would expect to find them in great abundance in this drug. But this is not so ; some of the samples gave no indication of the presence of alkaloids, and the others only afforded traces. The amount of nitrogen is lower in charas than any other hemp drug. The Gwa.lior sample yielded 1.75 per cent, of nitrogen, which apparently was derived from the vegetable matter, as the " mashak" Of Amballa, which contained half the amount of insoluble vegetable matter, gave half the percentage of nitrogen (0-81 per cent).

Remarks on majum.—The confection of hemp known as majum is differently made in various parts of India. In the north it is made with white sugar, in the south it is made with brown sugar and mixed with so many spices and other ingredients that it'is quite black.

Allahabad majum was a whitish soft mass of peculiar ghee-like odour, and containing vegetable debris similar to that from ganja. It contained 4 per cent. of ghee and I.3 per cent. of vegetable matter insoluble in water. There was no trace of alkaloid present.

Benares majum was in cakes of the same colour and odour, and was nearly pure sugar. It contained 1-4 per cent, of fat, and 1 per cent. of leafy portions ; no foreign seeds were discovered and no alkaloid could be detected.

Bombay majum was in the form of yellow-coloured cakes flavoured with spice, and with very little ganja. It contained 4-9 per cent, of ghee soluble in ether, and 21 per cent. of brown vegetable matter referable to cloves, cassia and saffron. No alkaloid was present.

Amballa ma/urn was in opaque white masses with a slight greenish tinge and odour of ghee. It contained 7.3 per cent. of ghee soluble in ether, and no vegetable structures of a green colour as in other samples. No alkaloid was found.

Hyderabad ma/urn was in the form of square cakes covered on one side with silvered paper. One kind was white and contained 2.3 per cent. of ghee, the other was coloured yellow with saffron and contained only 1.2 per cent. No seeds or vegetable structures were observed. This had been doubtless straihed through a cloth before being boiled down. With these samples came a mixture, wrapped up in a leaf, of powdered raisins and white sugar. When this was examined, several other substances were found, such as pieces of almonds, seeds of various kinds, stamens of some plant, some cereal grains and spice. The little black seeds were further examined microscopically and were found to consist of two kinds, one with three angles, probably from a polygonaceous plant, and the others were, in very small quantity, the seeds of dhatura. This is the only sample in which I have found dhatura seed present.

Caimbatore, Oolacamund and Tanjore nzajums were black soft masses of a liquorice-like odour and sweet taste. An analysis was not attempted of either sample, but they were each broken down with water, and amone.p the insoluble portion a search was made for dhatura seeds, flux vomica, etc., but without success. The amount of leafy organs, such as would come from the ganja, was in a very small proportion. In Ootacamund there is a black majum used for adults, and a white kind given to children, but as sold in the gan ja-shopthey are not very powerful preparations.

• Kaniyambadi, North Arcot, a villaf,e not far from Vellore, vhere some of the dealers in Javadi genie live.—(C. Beaten)