In 1942 Theorell (1) published a paper on the magnetic properties of crystalline horseradish peroxidase and some of its derivatives, in which he stated, These determinations are to be regarded as a stage in our attempts to throw some light on the general problem: What conditions the great differences in the effects of the hemin ferments, despite their apparent uniformity in the composition hemin plus protein? Since in all cases the fermentative reactions take place at the iron atom or atoms, one imagines that the way in which the iron atoms are built into the molecule is of decisive importance for their mode of functioning .... We must of course also take into account an interaction between the iron atom and more re mote groups in the protein part. This interaction must, however, be weak in rela tion to the effect of the groups in the protein part that are directly coupled to the iron atom. Our first task is thus to try to ascertain the way in which the hemin is bound to the protein part. But not even if this proves successful can one count on getting a full insight into the way a hemin ferment takes effect.In discussing his now classic information about effects of ligand differences on magnetic susceptibilities Theorell continued, Now that sufficient experience of the magnetic properties of the proto-hemin proteides has been gained, it is possible to establish ... rules regarding the relation between the spectrum and the way the iron is bound.He listed the spectral characteristics of the four classes of compounds known at that time: (a) ferric compounds with essentially ionic bonds; (b) ferric compounds with essentially covalent bonds: (c) ferrous compounds with essentially ionic bonds; and (d) ferrous compounds with essentially covalent bonds. He added the reservation that groups (b) and (d) cannot be unambiguously distinguished by absorption spectrum only. Conventional wisdom today modifies Theorell's classification by the use of the terms high-spin and low-spin for ionic and covalent, respectively, but the same problems posed in 1942 by Theorell-the relationships between structure, 1 The survey of literature pertaining to this review was concluded in October 1966 .2 Myoglobin and hemoglobin denote iron (II) forms, in preference to the terms deoxymyoglobin and deoxyhemoglobin. Metmyoglobin and methemoglobin denote the "acid" iron (III) forms; other derivatives will be clearly designated.