Visitors to animal houses in the U K are confronted with endless rows of rat cages and mouse boxes and it is very rare indeed that even such well-known animals as the Chinese hamster or the mult imammate rat are encountered. The question arises whether the impression gained holds true in other countries, or whether British animal houses show a nonrepresentative sample of experimental animals. Unfortunately, it is almost impossible to answer this question. Animal experinaents are carried out in many countries under greatly varying conditions, and in most of them, even in such scientifically minded ones as the USA and Japan, there is no central control over the work involving laboratory animals. In Britain, research workers are compelled, under the Act of 1876, to submit to the Home Office a yearly return of all animals used. These returns could form a useful statistical basis, but, under the law mentioned, there are only 3 groups of animals: Horses, asses and mules form one; cats and dogs the second; and all other species the third. Consequently, no relevant data could be obtained from these yearly returns, and it became necessary to look for other references. In the end, the UFAW Handbook (Worden & Lane-Petter, 1957) and the 6th supplement of the Federation Proceedings (Cass, Campbell & Lange, 1960) were chosen as the most suitable sources of information, and Tables 1 and 2 summarize the findings. In Table 1 a comparison is made between the number of pages dedicated in the UFAW Handbook (Worden & Lane-Petter, 1957) to and animals, and Table 2 analyses the number of references made to various groups of animals in the Federation ProceedhTgs (Cass et al. 1960). In making this analysis it was necessary to draw an arbitrary line between common and uncommon animals. Animals that are abundant in one country might be almost unobtainable in another, and obviously the final decision must depend on the circumstances under which one works. It was also necessary to introduce some generalizations into Table 2. The term rodent applies only to rats, irfice, guinea-pigs and Syrian hamsters; all remaining rodents have been included under others. Farm animals and non-mammals have been omitted altogether, and so have papers dealing with more than one animal species. Both tables confirm the original impression that there is a heavy bias towards the more common species, but there remains the question why such a bias should exist. Undoubtedly, one reason for this trend is the necessity for every scientist to produce results. Sooner or later, promotion and reputation depend on the number of papers published, and the temptation is very great indeed to follow the well trodden path and confine one's experiments to conventional animals. However, there might be a more deeply rooted reason for this conservative approach. The origin of medicine is lost in antiquity, but the experimental approach is comparatively new (Singer & Ur.derwood, 1962). One of the earliest references is to the work of Swammerdam (1637-1680) who introduced the modern nerve muscle preparation, using the frog as