A MICROSCOPICAL STUDY OF HONEY POLLEN.

INTRODUCTION.

METHODS OF EXAMINATION.

Little or no work appears to have been done hitherto upon the microscopy of American honeys, and the few records of the work which has been done upon European honeys were not available for consultation during the present study. It was therefore necessary to work out methods as well as results. As the structures found in honey are usually too thinly scattered for convenient inspection, it was found advantageous to dissolve the sample in several times its volume of water and centrifuge, and then examine the sediment with the microscope.

Preliminary to the examination of honey sediments, and also in connection with this investigation, it was found necessary to examine pollen taken directly from flowers for the purpose of comparison with that in the sediments. Either fresh flowers or herbarium specimens may be used. In the latter case, especially if the specimen be old, the pollen grains are apt to be so dried out and shrunken that it is' difficult to determine what their normal appearance really is. If the flower be moistened with ether, the pollen grains will return to their normal size and form. Some difficulty was experienced in finding a satisfactory mounting medium. Water has a limited application, but does not clear the pollen sufficiently. Solutions of chloral hydrate or of alcohol cause the grains to burst or become distorted. Strong solutions of glycerin have the same effect, but a 10 per cent solution produces no change in form, except that it sometimes causes the protuberances to swell and, in freshly gathered pollen, it may induce the pollen tube to germinate. Glycerin gives good results as a clearing agent, and if the cover glass is sealed with a suitable cement a fairly satisfactory permanent mount is obtained. The pollen grains found in honey were usually beautifully clear and transparent, and this suggested the use of a strong solution of glucose as a medium for mounting temporary slides. The solution has to be made up frequently, as it soon becomes filled with yeast cells, and for the same reason it is not suited for making per

manent slides. A considerable amount of oily matter is found sticking to the pollen grains of some plants which interferes with the clearing. This may be removed by placing the pollen, or the flowers containing it if they are small, on several thicknesses of filter paper and adding ether slowly. The ether dissolves the oil and passing on is absorbed by the filter paper.

STRUCTURES PRESENT IN HONEYS.

In addition to the structures normally found in honey, accidental contaminations are nearly always present. Whenever honey is exposed to the air for any length of time it is sure to collect more or less dust, which varies in character according to the conditions under which the honey was exposed. For example, the sediments frequently contained starch granules, and these were, moreover, of the kinds of starch which were being used or examined in the laboratory at the time that the honey was examined. The same fact may also account for the frequent occurrence of coniferous pollen, as a considerable quantity of pine pollen was collected and brought to the laboratory at this time. It is apparent, therefore, that care must be taken to expose the samples as little as possible, especially when examining honeys for the detection of adulteration. In honeys which have been adulterated with glucose, starch grains, frequently changed by heat, often occur, and these may have been added in the adulterant.

Crystals of various kinds are sometimes found in honey sediments. These include needle-shaped crystals and regular octahedrons and prisms identical in appearance with the calcium oxalate and calcium. phosphate crystals found in urine. Crystals of dextrose, which are frequently found in large numbers in honey, were usually dissolved in preparing for centrifuging, so that they did not appear with the sediment.

Structures of animal origin are occasionally found. These include hairs and other appendages of insects, scales of Lepidoptera, and sometimes muscle fibers and fragments of tracheæ.

Structures of vegetable origin are always present. They consist of hairs and other tissue fragments, fungus spores, and pollen grains, of which the last are the only ones to occur constantly and in sufficient numbers to be of importance.

Small fragments of the comb are frequently found in extracted honeys, but are usually easy to recognize. If, however, the honey has been heated sufficiently to melt the wax, the latter collects in spherical masses which may be confused with pollen grains, from which they may be distinguished by their lack of uniformity.

NUMBER OF POLLEN GRAINS IN HONEY.

The first work undertaken in the present study had for its object to find out whether it were possible, by determining the number of pollen grains in a definite quantity of honey, to judge of the adulteration of the sample. It is obvious that if the numbers present in a large number of samples of pure honey were found to be fairly constant, any large diminution of the number would indicate adulteration. The method used was as follows: One gram of honey was weighed out into a graduated centrifuge tube, taking care to get as nearly an average sample as possible. Distilled water was added, and the tube shaken until the honey was completely dissolved. After centrifuging for about five minutes the supernatant liquid was drawn off with a siphon, leaving less than 1 cc, and water was added carefully to bring the volume exactly to 1 cc. After thoroughly mixing this water with the sediment, a drop of the liquid was placed in a Thoma blood-counting chamber for examination. As the ruled area proved to be too small the volume of the entire chamber was found accurately, and all pollen grains in the chamber were counted. In the formula N N equals the number of pollen grains in the 1 cc of liquid in the centrifuge tube, or the number of pollen grains in 1 gram of the honey, equals the number counted in the Thoma chamber, and the volume of the chamber taken as a fraction of 1 cc. In actual practice ten counts were made on each sample, and the average taken as n. Duplicate counts were made in most cases by centrifuging a second sample and proceeding as with the first. The following table shows the results obtained:

n

From an inspection of these data it will be seen that the estimated number of pollen grains in 1 gram of the honeys examined varies between 123 and 5,410, and that a honey like No. 5 or No. 92 could be mixed with several times its volume of glucose without making the pollen count suspiciously low.

EXAMINATION OF POLLEN FROM FLOWERS COMMONLY VISITED BY BEES.

With the cooperation of Dr. J. N. Rose, of the National Museum; Dr. C. F. Wheeler, of the Bureau of Plant Industry, and others, the pollen of a large number of flowers known or suspected to be visited by bees in different sections of the country was obtained from authentic sources and used for comparison with that found in the honeys. As the pollen of different plants more or less closely related is often found to be very similar or even identical in appearance, it is frequently impossible to carry the identification as far as the species, and for this reason it has been thought best in most cases to refer the pollen to various types, subdivided if need be, which might, for the sake of convenience, be named after the family or genus represented. The descriptive data are therefore arranged alphabetically by types with the exception of the leguminous type, which on account of its importance is considered first:

Leguminous type.-On account of the importance of leguminous plants in the production of honey, the pollen was quite thoroughly studied. The typical shape is ellipsoidal when seen from the side. Each grain has three openings arranged at equal intervals around the equator, and usually indicated by rather large protuberances. The grains are light colored, with very finely granular contents, and thin, usually smooth walls. Longitudinal grooves are often seen as faint lines running in either direction from the openings.

a

The pollen of white clover (Trifolium repens L.) may be taken as typical. The size varies from 24 μ to 28 μ in length and 20 μ to 24 μ in diameter. Alsike clover (T. hybridum L.) pollen (Pl. II, fig. 1) differs from that of white clover only in the position of the openings, which are usually nearer one end of the grain. Red clover (T. pratense L.) pollen (Pl. II, fig. 2) is of the typical shape, but has an average size of 38 μ × 43 μ. The walls are rather thick and have a slightly reticulate appearance. Alfalfa (Medicago sativa L.) pollen (Pl. II, fig. 3) differs from the last only in the smooth and thinner wall. Sweet clover (Melilotus sp.) has pollen which can scarcely be distinguished from that of white clover. It is, however, a trifle longer in proportion to the diameter, averaging about 30 μ × 24 μ. The pollen of honey locust (Gleditsia triacanthos L.) is like that of red clover, but somewhat more rounded. Locust (Robinia pseudacacia L.) pollen is roundish or even flattened so as usually to be seen on end. It has a diameter of 30 μ to 35 μ. Mesquit pollen (Pl. II,

The unit employed in microscopical measurements is the micro-millimeter or micron, equal to ro of a millimeter or about 3 of an inch. It is usually indicated by the Greek letter Mu (μ) placed after the number.

μ.

fig. 4) or algarroba a (Prosopis juliflora (Swz.) DC.) is of similar shape and has an average diameter of 32 μ. Yellow wood (Cladrastis lutea (Michx.) Koch or C. tinctoria Raf.) has very small pollen grains (12 μ × 14 μ to 16 μ) of the usual leguminous form. The pollen from the species of Acacia studied differs remarkably from that of other leguminous plants examined. The grains are compound, each consisting of 16 granules. In cat claw (Acacia wrightii Benth.) these grains have a diameter of 35 μ to 40 μ and a thickness of 24 μ, while each granule has a diameter of 10 μ (Pl. II, fig. 5).

Basswood or linden type.-The pollen of basswood (Tilia americana L.) is quite distinctive. The shape is somewhat lenticular, so as nearly always to present the same aspect. The appearance is that of a regular three-lobed figure inscribed in a circle. The three openings are subtended by the clear spaces between the lobes. The diameter averages 32 μ (Pl. II, fig. 6).

Balsam type.-Impatiens sp. has oblong, light-colored, thin-walled pollen grains measuring 20 X 24 μ to 28 μ. The four openings are located at the corners. There are a good many needle-shaped crystals in the anther tissues (Pl. III, fig. 1).

Cactus type.-Opuntia sp. has pollen grains 125 μ in diameter of approximately spherical shape (Pl. III, fig. 2). They have a number of faces, each provided with an opening, and the intervening space is reticulated. Ariocarpus sp. has spherical or subtriangular pollen, 65 μ in diameter, with a reticulated surface and only three openings.

Citrus type.-Orange and lemon pollen (Pl. III, fig. 3) are practically alike. The grains have a finely indented surface which gives them a rough appearance. There are normally four or five openings. In the samples examined the grains varied considerably in size, many appearing not to be well developed. The average size of the better developed grains was found to be 30 p to 40 p.

μ

Composite type.-There is but little difference in the pollen of the various species of compositæ. It is usually rather small, spherical, provided with three openings, and covered with spines. Aster pollen (Pl. III, fig. 4) is thickly covered with short, sharp spines. The grains are 16 μ to 20 μ in diameter. Golden-rod (Solidago sp.) pollen is very similar and averages about 20 μ in diameter. The pollen of Pluchea sericea (Nutt.) Coville (P. borealis Gray) is 24 μ to 28 μ in diameter, with fewer, longer, and stouter spines. Bidens pollen is 24 in diameter and has spines a little more slender than those of the Pluchea. Sunflower (Helianthus sp.) pollen is 35 μ to 40 p in diameter and rather thinly covered with sharp spines. In ragweed

μ

μ

μ

a Hawaiian algarroba=Prosopis juliflora; Texas mesquit=P. glandulosa Torr.; Arizona mesquit=Prosopis velutina Wooton.; European algarroba Ceratonia siliqua L. The mesquit pollen in the Texas and Arizona honeys was not identified as to species.