Many important taxonomic characters are difficult or impossible to observe with the naked eye. For example, you may wish to count the number of tiny teeth at the apex of a ray floret of a composite, or determine whether a leaf has minute hairs in the axils of the veins. Anatomical structure of parts such as roots and bark is particularly hard to see. A good hand lens, used in good light, will suffice for many purposes. However, for those who will be looking at a great deal of material, a dissecting microscope is very useful. This should not be confused with a light microscope, which has a built-in light in the base, a raised stage for slides, and several objective lenses for increased magnification. A dissecting microscope has a much lower range of magnification and a low stage on which a fairly large object can be placed, with room under the head of the microscope for hands holding tweezers, etc. A separate lamp illuminates the object. Dissecting microscopes are much more comfortable to use than hand lenses, with binocular eyepieces and a greater maximum magnification that provides better resolution of very small features. Forceps and perhaps dissecting needles should also be obtained, as they allow convenient manipulation of small plant parts.

One must expect harvested plant material, even of a whole herb, to look very different from live plants of the same species. Unless plants are small, people do not normally collect whole plants. Larger plants would typically be cut or broken into pieces; only the tops may be collected, or lower leaves may be stripped from thick stems, which are then discarded. Since dried material is often brittle, more breakage can occur during initial processing. Some characters, such as the height of a large plant, usually cannot be observed, and others may be observed only imperfectly. For example, if a plant has large, petiolate lower leaves but smaller, sessile upper leaves, broken pieces with each type of leaf base may be found, but not attached to a stem in a neat sequence.

Dried material also may change appearance greatly by shriveling, which can make parts hard to measure, and colors may change to some extent (although if colors are too faded, it is often an indication of poor quality due to improper processing). Dried flowers may lose parts such as petals, which may be found loose. Most dried leaves, flowers, and fleshy fruits are more easily examined if they are rehydrated. Various wetting solutions may be used to rehydrate herbarium material. The simplest is to use hot water with a very tiny spot of dish soap in a watch glass or similar container. Solid or hard fruits do not rehydrate as quickly, and the use of boiling water can be helpful. With a few chemicals, a more efficient wetting solution can be made. Dr. Lynn Clark’s Modified Pohl’s Solution (based on a recipe by the late Dr. Richard Pohl) uses 750 ml distilled water, 250 ml 1-propanol, and 2 ml Ivory liquid soap to make one liter of solution; it can be kept in dropper bottles. Once material has softened, it can be spread out with forceps so that its original shape and size can be seen.

Sometimes the initial processing of an herb leaves it broken into rather small pieces, which is less than ideal, but informative parts may be found with patience and care. Small flowers often remain intact in broken material, although inflorescences usually do not. Information regarding the likely shapes of the leaf base, apex, or margins, the venation pattern and the presence of surface pubescence or glands may be obtained from broken leaves by sorting through the pieces to find those that include suitable portions of the leaf. If leaves are broken, it may or may not be possible to estimate their size accurately by examining the larger pieces. Leaf size is more likely to be an important character in plants that have uniformly small leaves, rather than those that have leaves potentially ranging in size from small to enormous. Small leaves, even when broken, often include enough of the leaf in a single piece to allow some estimate of the size to be made. (The exception is if material has been comminuted, or broken into tiny fragments in which few taxonomic characters remain visible. A few plants are traditionally treated like this during their initial processing, but botanicals of Western origin ought to have their identity certified before they are processed to that degree.) Sliced roots and rhizomes should be sorted through to find pieces that give the best view of the internal structure, which means those that are cut most nearly straight across rather than at an angle.

Fortunately, quality control of botanicals does not usually involve actual identification, but confirmation of identity. One doesn’t start with an unlabeled unknown and have to determine from scratch which of the 300,000 or so species of higher plants it is, which can be an exhausting task even with the best of material. Rather, one starts with a plant that has already been named by someone else, and must determine only whether that initial identification seems to be correct. If not, one simply rejects the material, without any need to figure out what it actually is. Knowing what the plant is supposed to be makes the use of botanical literature (including this manual) much simpler. One only has to look up descriptions of the plant named and ask: are these consistent with the material at hand, or not? The material can also be compared to illustrations, or to previously obtained reference plant materials of known identity. Such comparisons are not always straightforward: some plant species have highly variable leaf shape or pubescence, for example, and material that does not look much like a given illustration or reference sample might in fact be the same species. Written descriptions can make it clear when features of a species are variable, thus avoiding “false negatives” in which correctly identified material is wrongly rejected.

For those who are buying a large amount of a single botanical, a sampling scheme is necessary to ensure that a sufficient amount of the material has been checked for identity. Otherwise, one might by happenstance check the only portion of the material that did not contain excessive levels of some adulterant! A useful approach is outlined by the World Health Organization’s manual, Quality Control Methods for Medicinal Plant Materials (1998). In WHO’s recommended system, at least 10% of the containers in a batch should be sampled, rounded up, with a minimum of five containers (or all of the containers if the batch includes four or fewer). For each container, equal samples should be taken from the upper, middle and lower portions of the material, mixed, then repeatedly mixed, quartered, and divided in half by returning diagonally opposite portions to the container, until a suitable amount remains. These samples would also be used for any laboratory tests performed, such as ash content and screening for chemical markers, and for preservation of voucher specimens for each batch in case doubts about its identity should arise later.

Comparing a written description to illustrations or to reference specimens often enhances one’s mental picture of a plant. Previously obtained reference material of known identity is particularly useful because it can convey information on color, texture, and the like in more detail than any description or illustration. To avoid “false negative” results in which the identity of material is wrongly called into question, one must remember that an illustration or a piece of reference material represents a single individual, and thus can display only a portion of the variation that is present within a species. For example, if the flower portrayed is hairless and the flowers of the material to be identified are hairy, they may belong to the same species, yet appear substantially different at first glance. In case of apparent discrepancy between the material examined and the illustration or reference material, one must therefore consult a written description to determine whether the suspect character is indeed found in that species.