To obtain and enunciate correctly the problems suitable for discussion in each branch of science, you must have before you tables of dissection and logical division, and take them as guides;[52] beginning with the highest genus and proceeding downward through the successively descending scale of sub-genera and species. If you are studying animals, you first collect the predicates belonging to all animals; you then take the highest subdivision of the genus animal, such as bird, and you collect the predicates belonging to all birds; and so on to the next in the descending scale. You will be able to show cause why any of these predicates must belong to the man Sokrates, or to the horse Bukephalus; because it belongs to the genus animal, which includes man and horse. Animal will be the middle term in the demonstration.[53] This example is taken from the class-terms current in vulgar speech. But you must not confine yourself to these; you must look out for new classes, bound together by the possession of some common attribute, yet not usually talked of as classes, and you must see whether other attributes can be found constantly conjoined therewith. Thus you find that all animals having horns, have also a structure of stomach fit for rumination, and teeth upon one jaw only. You know, therefore, what is the cause that oxen and sheep have a structure of stomach fit for rumination. It is because they have horns. Having-horns is the middle term of the demonstration.[54] Cases may also be found in which several objects possess no common nature or attribute to bind them into a class, but are yet linked together, by analogy, in different ways, to one and the same common term.[55] Some predicates will be found to accompany constantly this analogy, or to belong to all the objects _quâ_ analogous, just as if they had one and the same class-nature. Demonstration may be applied to these, as to the former cases. [Footnote 52: Analyt. Post. II. xiv. p. 98**, a. 1. [Greek: pro\s de\ to\ e)/chein ta\ problê/mata, le/gein dei= ta/s te _a)natoma\s_ kai\ ta\s diaire/seis, ou(/tô de\ diale/gein, u(pothe/menon to\ ge/nos to\ koino\n a(pa/ntôn.] This is Waitz's text, which differs from Julius Pacius and from Firmin Didot. Themistius (pp. 94-95) explains [Greek: ta\s a)natoma\s] to be anatomical drawings or exercises prepared by Aristotle for teaching: [Greek: kai\ ta\s a)natoma\s e)/chein dei= prochei/rôs, o(/sai pepoi/êntai A)ristote/lei]. The collection of Problems or questions for investigation was much prosecuted, not merely by Aristotle but by Theophrastus (Schol. p. 249, a. 12, Br.).] [Footnote 53: Analyt. Post. II. xiv. p. 98, a. 5-12.] [Footnote 54: Ibid. a. 13-19. Aristotle assumes that the material which ought to have served for the upper teeth, is appropriated by Nature for the formation of horns.]