One of the wearisome problems of using the body of scientific literature to delve into specific ferret problems is the limited nature of that literature. While ferret information is rich in those areas used to model human and canine disease, it is frustratingly sparse in those areas that mean the most to ferret owners (this is gradually changing). While you can use the literature to discover a tremendous amount of information regarding influenza in ferrets (done to model the disease in humans), or on distemper (done to model the disease in dogs), finding pertinent information about nutrition is rather restricted. As mentioned, ferrets have been used to model disease in humans and canines. When an animal is used in this regard, it is a specific type of modeling called either an analog or a homolog. In the simplest biological definition, a homolog is similar structures with a common ancestry, like a human hand and arm, the front flipper of a whale, a bat wing, and a ferret front leg. An analog is similar structures without a common ancestry, such as the wings from a butterfly and hummingbird. These definitions are simplistic in terms of modeling disease because although ferrets and humans are mammals with a homologous physiology, they are not exactly the same. Mustelids diverged from the common mammalian lineage nearly 18 million years ago, and differences between the two have been slowly introduced since that time. It is clear that two structures or physiologies may, in fact, be biologically homologous, but enough differences may exist to shift them back to analogs. For example, dogs and humans share homologous physiologies, but dogs cannot eat Tylenol or chocolate. So, while the physiology between the two species are homologs in a general biological sense, in a specific physiological sense they are more like analogs. The problem was addressed by Hodder, who in 1982 distinguished FORMAL from RELATIONAL analogies, arguing formal analogs were far weaker models. Simplistically, the simplest way to distinguish between the analogies is that in a formal analogy the similarities are inferred, while in a relational analogy the similarities are observed (the difference is far more complex, but beyond the scope of this discussion). A dog's inability to eat Tylenol or chocolate while a human can, demonstrates the two are formal analogs and are not relational. This is why the human ability to metabolize chocolate cannot be used to directly model canine theobromine metabolism. When ferrets are used to model canine distemper in dogs, the observable similarities support the use of a relational analog. This means the data generated from ferret distemper studies can be used to generate hypotheses regarding the disease in canines, as well as using the disease in dogs to model distemper in ferrets. In a relational analogy, data can be used in both directions. It also means that if a ferret is a relational analog for a disease in dogs, and dogs are a relational analog a disease in, say, raccoons, then raccoons and ferrets can be used to model the disease in the other. Scientists are skeptical, conservative bastards, so a confirming experiment is desired, but not necessary for the creation of hypotheses. This is critically important because it directly affects the use of other species to fill in gaps in our knowledge of ferrets, as well as how we use and interpret hypotheses. There is a lot of misunderstanding regarding what a hypothesis is, and how it works. A hypothesis is nothing more than an idea or proposition that is based on certain defined assumptions that can be evaluated scientifically. What that means is the hypothesis itself IS NOT tested, but the assumptions it is based upon are confirmed or falsified. If the assumptions are falsified, that is, proven wrong, then the hypothesis is rejected. Bob C [Posted in FML issue 3949]