Billion Confusion "the molar of X is 25,000 g In Vivo Difficulties mol-I", "the of protein SIR,-The use of the terms "in vitro" and SIR,-Teodor Juskiewicz (Nature, 228, X is 25,000 daltons", and "the relative "in vivo" is now deeply entrenched in 297; 1970) referred to the American molecular mass (that is, molecular ) the scientific literature. They are used practice of using the word "billion" to of protein X is 25,000".) to denote the difference between experi­ mean 109 instead of 10 12 and he appealed The 14th General Conference of ments performed outside the living to American colleagues not to use the and Measures (CGPM) of the organism (although often with living misleading term parts per billion. International Bureau of Weights and tissue, and those carried out inside the I support Juskiewicz's appeal and Measures, convening in 1971, will con­ organism. The use of the two terms, suggest that it is that some agree­ sider a recommendation approved in 1969 although hallowed by time, sometimes ment was reached to avoid misunder­ by the International Committee on causes difficulty, especially as editors of standings, which can arise by the use of Weights and Measures (CIPM) to include learned journals differ in the extent of this, at present, equivocal word. the as a of the International their tolerance and of their To my mind the word "billion" means System of Units (SI), besides the six base pedanticism. Some editors are prepared a million to the of two, similarly units on which the system was established to accept the terms virtually as the author "trillion" means a million to the power of in 1960 (the , the , the proposes, irrespective of syntactical or three and so on using suitable prefixes , the , the , and the scientific niceties. Some turn a blind eye added to the root "-illion" for numbers of ). The additional base unit is to their suspiciously foreign sound and the type 106n (where n is an integer). needed to introduce SI units for the are prepared to admit them as current There is some need for a simple name "molar" physical quantities (molar English usage. Others, more severe, by for the number 109 which would be , , molar capa­ clapping the terms in itiilics, clearly still preferable to the rather clumsy "thousand cities, molar of formation, etc.). regard them as aliens against whom the million". The word "milliard", obviously The appropriate corre­ innocent reader must be warned. familiar to Juskiewicz and, I understand, sponding to the concept that different Hyphens between the two parts of each currently used in France, seems an substances have natural molecular consti­ term are not usually required, but on obvious choice. Furthermore, this word tutions (the word "molecular" here being occasions have been insisted upon by could form the basis of a system of used in a broad sense to include any the jlJiberal of outlook. Although naming of the type specified constituent entities, whether strictly "in vitro" and "in vivo" are 1O(6n+3) in the same way that "million" they be , , , pairs, adverbial phrases (and are only so used has for the 106n numbers. Thus 1015 or other aggregates) has not until recently by cognoscenti) they are now often would be called a billiard and 1021 a been identified by a commonly recognized misused as adjectives. Hence one reads trilliard, and so on. name. The name, "amount of sub­ of "in vitro experiments" and the even stance", has now been adopted by the more disgraceful "in vitro results". Even Yours faithfully, International Union of Pure and Applied "semi-in-vivo" (hyphenated surely) ex­ , the International Union of periments have recently been threatened. . M. BOROUGHS Pure and Applied , and the Fowler, unfortunately, wrote before such International Organization for Standard­ indignities became common, and has 55 Wolsey Drive, ization to define a physical quantity nothing to say on the matter. Walton-on- Thames, proportional to the number of constituent With due awareness of Lord Chester­ Surrey entities of that substance (molecules or field's famous maxim, I should like to other entities, such as may be specified by suggest the introduction of two new a ). The proportional­ words to replace "in vitro" and "in vivo". ity factor is the same for all substances These would be the Simple adjectives Molecular Mass and may be taken to be the reciprocal of "vitral" and "vival" respectively. The the . A unit for the SIR -Dr Edsall has explained (Nature, new words are short, their derivation is physical quantity, the mole, has long been 22S', 888; 1970) the useful distinctions etymologically pure and their meaning by recognized. The definition given by the past association) is clear, when referring that should be preserved among the CIPM in 1967, confirmed in 1969, and expressions, molecular mass, relative either to the tests themselves or to the included in the draft proposal prepared molecular mass (commonly called "mole­ results of tests. Moreover, they will for the 14th CGPM introducing fhe mole never need italics or hyphens. Reference cular weight") and molar mass. These as a base unit in the SI, is as follows I : to various technical dictionaries has dis­ quantities have respective dimensions: mass unity ("dimensionless") and mass The mol is the of a closed, perhaps surprisingly, no prior use , I X (amount ofsubstance)- . Thecommon system which contains as .many eleI?entary of such adjectives. Although their intro­ entities as there are atoms m 0.012 kilogram duction, it is realized, would not be so unit of molar mass (not its dimension) is of 12. traumatic as the introduction of SI units, the per mole (symbol, g/mol or nor so subversive as the substitution of g mol-I). Among recognized units of Note: When the mol is used, the elemen­ tary entities must be specified and may be "retinol" for vitamin A', there are bound molecular mass is the unified atoms, molecules, ions, , atom of the nUclid.e 12C (l u= 1.660 53 x 10 - 27 kg approxI­ If the 14th CGPM accepts the mole so Yours faithfully, mately), and for which Dr Edsall r~om­ defined as' an SI base unit, then the SI unit of molar mass will be the kilogram J. mends the simpler name, , WIdely per mole (kg mol- '). This unit is large 19 Vineyard Hill Road, used by biochemists. (His examples of for ordinary chemical purposes and the Wimbledon, different statements expressing the same common unit, gram per mole (l g mo\-I London SW19 fundamental facts should have read:

© 1971 Nature Publishing Group NATURE VOl. 229 JANUARY 8 1971 143 = 10-3 kg mol-I), will continue in use as Definition of Intelligence axiomatize the levels of the system inde­ an accepted decimal sub-multiple of the pendently and use negative feedback to SI unit. However, the SI unit itself is SIR,-Because others (Nature, 228, 1008; control the universality (requisite variety) suitable for expressing values of the molar 1970) have commented on the defini­ of both individual levels and the system as mass for macromolecular substances. tion of intelligence put forward by a whole. An essential part of the control Thus, one could add to Dr Edsall's Fatmi and Young (Nature, 228, 97; 1970) system is an order : disorder detector, as statements the alternative: and subsequently extended by myself suggested by H. B. Barlow in his letter. "the molar mass of protein X is 25 kg (Nature, 228, 589; 1970), I would like to This type of system would also imply that mol-I". make some further observations. the original definition of Fatmi and Although the unit of mass, "unified With regard to the letter from P. M. Young would be too broad to draw a atomic mass unit", is outside the SI, it Muller, the process of induction would be valid distinction between men and has been recognized by the CIPM as covered by my own definition, as would machines. useful in specialized fields of scientific "synthetic a priori". However, pro­ With regard to the letter from H. A. research I . Its value expressed in the SI cesses of reasoning from the part to the Cook. Any system is quantifiable if one unit, the kilogram, is derived by experi­ whole, from the particular to the general, knows what to measure and how to ment and is therefore not known exactly. and from the individual to the universal, measure it. It is another matter to Although in general one should be chary are not identical and isomorphic pro­ decide whether such a quantification of proliferating special names, the present cesses, nor are they symmetrical with provides an adequate description of the name for this unit, even when contracted respect to deduction and induction. system as a whole. The information­ to "atomic mass unit" (the term "unified" If we accept Godel's theorem, a single, theoretic brain model mentioned above distinguishing it from slightly different finite automaton with a phrase-structure implies a physiological symbol-processing earlier units based on 160 and on 0= 16), grammar can be either complete and con­ mechanism in the brain, which could form is not notably convenient or informative. sistent (closed) or universal (open). A a substrate for the heritable components Dr Edsall's suggestion that it be renamed Godel complete system will accept only of both intelligence and linguistic be­ the dalton merits consideration by the tautologies or empirically verifiable sense­ haviour. Since such a component would international agencies concerned with data that have been specified in the be determined by input-rate and digit­ standardization of chemical and physical instruction set, rejecting all other inputs span, and would be invariant with respect nomenclature so long as the unit itself as illogicality or "noise". The possi­ to learning, it should be possible to continues to be recognized by the CIPM bility of new or unspecified states is isolate and quantify it given a calibrated as one of those useful outside the SI in excluded. If we assume that human digital input and criteria for assessing the specialized fields. It would not be helpful intelligence not only construes syllo­ output response in quantitative terms. if scientists in different fields employed gisms, but also discovers, elucidates and With regard to a quantified definition of different names for the same unit. initiates the previously unknown, then it intelligence, without recourse to linguistic is clear that this cannot be a property of behaviour, we may be up against a special Yours faithfully, Godel complete automata (existing com­ case of Richard's paradox. puters). Indeed, it would seem that an MARTIN A. PAUL The correspondence following Fatmi incomplete instruction set (a quasi-non­ and Young's original letter has empha­ Division of Chemistry and deterministic system) is a necessary con­ sized the need for further dialogue con­ Chemical Technology, dition for the emergence and evolution of cerning the theoretical and philosophical National Academy of Sciences­ intelligence, as we observe it in living aspects of machine and human intelli­ National Research Council, systems. However, the possibilities of gence, and it is gratifying to learn that the Washington, DC 20418, USA machines do not end with single finite­ Cambridge Branch of the Brain Research state machines or with phrase-structure I The International System of Units (S/), Association is setting up a forum in this translation approved by the International grammars. area of study. Bureau of Weights and Measures of In my own work I have been con­ its publication, Le Systeme International sidering the possibilities of hierarchical Yours faithfully, d' Unites, prepared jointly and published independently by the National Physical networks of automata, some of them GORDON HYDE 1 Laboratory, UK, and the National Bureau of backward deterministic , in an attempt to 11 The Close, Standards, USA: National Bureau of solve the Godel theorem problem for Dunmow, Essex CM6 1 EW Standards Special Publication 330, 1970, quasi-non-deterministic systems, includ­ US Government Printing Office, Washing­ I Wang How, A Survey of Mathematical ton, DC 20402. ing brain-models. The basic idea is to Logic, 175 (North-Holland, 1964). Obituary end of the war returned to research in the assistance from the National Health Dr J. E. Falk university on a Wellcome fellowship and Medical Research Council he went JOHN EDWIN F ALK, chief of the under Professor Adrian Albert. During to University College Hospital Medical division of plant industry, Common­ this period his research into the mode of School in London to work with Professor wealth Scientific and Industrial Research action of certain antimalarial drugs C. Rimington, FRS. His research in­ Organization, Canberra, died on October awakened his interest in biochemistry and cluded a systematic study of the infra­ 25, 1970. Born in 1917 at Cessnock, led him into the field of haem pigments. red spectra of porphyrins and haems as NSW, he studied pharmacy at the In 1947 he was awarded a grant by the well as further work on cytochrome University of Sydney, but finding that National Health and Medical Research oxidase. He received his PhD in 1951. his interests lay in chemistry he com­ Council to work at the Royal North A "Nuffield Unit of Pyrrole Pigment pleted a Bachelor of Science degree in Shore Hospital, Sydney, under Professor Research" was established in Professor 1942. After graduating he joined Profes­ M. R. Lemberg, FRS. Here he investi­ Rimington's department and Falk was sor V. M. Trikojus in an investigation of gated the prosthetic group of cytochrome appointed director. In 1953, he was synthetic methods of preparation of oxidase and learnt a great deal about awarded a Foulerton research fellow­ essential drugs unavailable in war porphyrins. ship by the Royal Society. During this time Australia. In 1944 he became Chief In 1948, Falk was granted a Nuffield period, he developed analytical methods at Bayer's in Sydney and at the travelling fellowship and with financial for porphyrins and turned his attention

© 1971 Nature Publishing Group