The reasons why particular substances were selected as homeopathic remedies are complex and varied. Many were familiar from traditional Western folklore or herbal lore. Some, such as mercury, were used in contemporary conventional medicine. Others included minerals or elements that had been used as nutritional supplements, such as zinc. Out of curiosity, or because they had a long herbal tradition or were known to have a strong, even toxic, effect, different substances were tried and the information cataloged. The greatest influence came initially from herbalists in Europe but, as knowledge grew of the medicinal traditions of other cultures, more substances were tested for their homeopathic potential. European explorers and settlers filtered back information amassed on their travels. Dr. Constantine Hering, for instance, visited South America and discovered the healing properties of the bushmaster snake.

Hahnemann set down strict guidelines for testing, or “proving,” potential remedies. This term developed from Prüfung, the German word for a trial. A set of volunteers, or “provers,” take a trial substance in different strengths, or potencies, and make a detailed, archetypal “character” emerges. Essential or “keynote” characteristics, both mental and physical, can be established to give a thumbnail summary, enabling homeopaths to recognize cases where a particular remedy is appropriate. Dr. James Tyler Kent carried out a great deal of research to enlarge knowledge of the remedies in the daily record of their moods, sensations, and any symptoms that develop. Symptoms are categorized as general, relating to a temperamental picture, or specific, affecting a particular part of the body. Surrounding or provoking circumstances, and triggers that make symptoms better or worse, are all noted. Certain general, physical, and psychological affinities appear more evident than others, and some symptoms will be very common among the provers; for example, a headache first thing in the morning that gets better after eating, or indifference to loved ones. These are “first-line” symptoms. Other symptoms may be experienced by only a few, or even one, prover. These are known respectively as “second-line” and “third-line” symptoms. Any symptom that can be shown to be unlike the person’s usual state of health will be recorded. This information is then compiled to produce a “symptom picture” that takes into account the potential variations produced by different provers’ constitutional types. Often this will then be compared with information about the substance and its characteristics, possibly from its prior use within herbal or folk medicine, or from knowledge of its toxicology, to deepen understanding of the remedy. As understanding of each remedy’s “symptom picture” grows, an materia medica, and his work has been built on and augmented by many other practitioners.

Beyond the individual remedies, it is possible to make connections between remedies and to establish group relationships. Studying the group of remedies based on potassium or calcium compounds, for example, reveals themes, such as physical weakness with all Kalium remedies, or sensitivity and shyness with all Calcium remedies. This thematic analysis is most obvious in clear categories like families of the periodic table, which have recently been researched by the British homeopath Jeremy Sherr. Relationships between plants within the same botanical family, or between animals with common links, can also be found. For example, all snake-based remedies tend to affect the blood and nervous system, and are for highly oversensitive individuals. The Indian homeopath Rajan Sankaran is working on an approach to case-taking that emphasizes the underlying sensations a patient feels, and linking this to a framework running through the animal, plant, and mineral kingdoms and families that form the basis of remedies. The breadth of information available from some of the old provings is not complete, whereas other provings are very well documented. New provings tend to be carefully managed and usually give a full picture.