March 17, 1H93.] 



SCIENCE. 



149 



or other small bird would be overwhelmed. As it is, the cow- 

 bird studies the limits of endurance in its victinns and rarely ex- 

 ceeds the bounds. The most egs;3 I ever found in a nest infested 

 by cow-birds was nine, and the species generally lays only two or 

 three eggs, thus generally keeping the outside limit to six or 

 seven, with the owner's eggs. 



THE USE OF THE TERM "CARBOHYDRATES." 



BY W. E. STONE, PH.D., PURDUE UNIVBKSITY, LAFAYETTE, IND. 



It has frequently happened in the history of chemistry that 

 names and terms have lost their original significance so soon as 

 the knowledge of the bodies tn which they were applied has be- 

 come more extended. " Organic" chemistry is better named the 

 " chemistry of carbon compounds;" the " ai'omatic " bodies have 

 disappeared in the broader designation of benzine derivatives. 

 In the same way it appears that we have reacht-d, or already 

 passed, a transition stage in the use of the term '• carbohydrates." 

 Ti-eatises on chemistry still retain the old definition of the term, 

 while those familiar with recent progress in this 8eld no longer 

 feel themselves restricted to these ancient limits. It is the pur- 

 pose of this paper to consider the present status of this subject. 



Von Lippmann. in his work "Die Zuckerarten und ihre Deri- 

 vate," adopts Fittig's view that the carbohydrates are derived 

 from the hypothetical heptatomic alcohol CoHj(OH),, which, by 

 lossof water, forms the simpleor complexanhydrides, CjHjoOj or 

 CjoHjgOji, known as sugars. His treatment ignores the ex- 

 istence of any carbohydrate with less than six carbon atoais, 

 although he says that, with the (at that time, 1882) slight knowl- 

 edge of the constitution of the carbohydrates, it was impossible 

 to regard this definition as final and complete. 



In 1888 appeared Tolleris' " Handbuch der Kohlenhydrate." in 

 which the definition of "carbohydrates" was limited strn'tl.v to 

 the bodies composed of C, H, and O, containing six carhon atoms, 

 or some multiple of six, and H and O in the same proportirm in 

 which they are found in water. But already Kiliani had shown 

 that arabinose, which had long been regarded as a true carbohy- 

 drate on account of all its reactions, had really the composition 

 CnHjoOj. Moreover, it had already been established that the 

 best known sugars, such as dextrose, levulose, galactose, and 

 arabinose, had the constitution of aldehydes or ketones of the 

 hexatomic, respectively pentatomic, alcohols. In anticipation, 

 therefore, of evident progress along this line, Tollens remarks in his 

 preface that such bodies as araliinose and the impending erythrose 

 might well be regarded as carbohydrates, but he retains the hex- 

 atomic nature as a requirement for the " true carbohydrate," and 

 puts all non-conforming but similar bodies under the head of "den 

 Kohlenhydraten nahestehenden Korper." 



Up to this time a sort of understanding had prevailed that the 

 carbohydrates were exclusively products of natural forces. It 

 had also been noted that these bodies gave certain reactions, which 

 were also presented as a basis for the classification given. 



These reactions, as stated by Tollens, are : — 



1. Reduction of alkaline metallic solutions. 



3. Rotation of polarized light. 



3. Subject to alcoholic fermentation by yeast. 



4. Formation of levulinic acid. 



5. Formation of characteristic compounds with phenylliy- 

 drazin. 



6. Certain color reactions. 



7. Solubility, either before or after hydrolysis. 



8. Decomposition by heat. 



AH of which hold strictly true for the hexatomic carbohydrates. 

 This classification was probably as liberal as the stale of knowl- 

 edge at that time would justify. 



But this classification is evidently arbitrary and ought not to 

 have weight in comparison with any classification based on 

 chemical constitution. If a similar constitution can be proven for 

 a series of bodies, the fact that they respond to certain reactions 

 will only be additional proof of their relationship. Such reactions 

 must, of course, be general in their nature, while special reactions 

 will only serve to characterize individuals. In this way the class 

 of carbohydrates must eventually include only bodies of certain 



constitution, while the characteristic reactions will be limited to 

 a smaller number, of more general application. A similar devel- 

 opment has taken place in the manner of classifying the hydro- 

 carbons, alcohols, acids, glycerides, etc. 



Of the carbohydrates conforming to the old definition, dextrose, 

 levulose galactose, and mannose are types. They respond to 

 the reactions given and have been found to possess the constitu- 

 tion of ketones or aldehydes of the hexavalent alcohol, C^Hj^O,.. 

 But we know two bodies of the formula CbH,„05, arabinose and 

 xylose, which are also aldehyde alcohols, and which give the 

 same reactions as their homologues, with the exception of fer- 

 mentation and the formation of levulinicacid. Again, we know 

 an aldehyde of the tetratomic alcohol erythrit, called erythrose. 

 of the formula C4H8O4, which responds to the same general re- 

 actions as its homologues. Glycerose, CgHiiOj, has also been 

 studied and found to correspond to the others of the series in con- 

 stitution and general reactions. It is even fermentable with yeast 

 like the regular carbohydrates, which shows this to be an inter- 

 mittent reaction when applied to an homologous series. Begin- 

 ning again with the group CoHisOj. we find that there have been 

 prepared synthetically three other homologues representing alde- 

 hydes, respectively of the hept-, oet- and nonatomic alcohols. 

 These also respond to the general reactions given, except that 

 they do not form levulinic acid. Heptose and octose do not fer- 

 ment, but nonose, with its multiple of three carbon atoms, is 

 fermentable. 



It is no argument against the carbohydrate nature of these 

 bodies to say that they do not occur in nature, since two of the 

 hexoses (galactose and mannose) have never been found free, but 

 are only known as derivatives of certain natural products. In 

 this respect they are on precisely the same footing as arabinose, 

 xylose, erythrose, and glycerose. 



It appears, therefore, that we have an homologous series of alde- 

 hyde or ketone alcohols of the general formula CnH^nOn with 

 these common properties : 1° sweet to the taste; 3° optically ac- 

 tive; 3° reducing alkaline metallic solutions; 4° yielding with 

 phenylhydrazin characteristic crystallinic compounds. Other re- 

 actions, such as great solubility, decomposition by heat, and color 

 reacti'ins, are less characteristic, although po-sessed in common. 

 Those containing three, or multiples of three, carhon atoms un- 

 dergo alcoholic fermentation with yeast, and this periodical re- 

 action seems an additional argument for their common nature. 

 Individually they yield, when heated with strong acids, charac- 

 teristic derivatives; for instance, the pentoses yield furfurol; the 

 hexoses levulinic acid ; others have not been carefully studied in 

 this direction. 



Following are the members of this homologous series which 

 are known, although several additional isomers are possible: — 



Triose, CsHeOj.— Glycerose. 



Tetrose. C4H3O4.— Erythrose. 



Pentose, C5H10O5. — Arabinose, xylose. 



Hexose, CsHi^Os.— Dextrose, levulose, galactose, mannose, 

 all in isomeric forms. 



Heptose, C^H^O,. — Heptose. 



Octose, CgHijOg. — Octose. 



Nonose, C9H13O9. — Nonose. 



By the definition of carbohydrates, now extant, only the hex- 

 oses are included. It is the purpose of this paper to propose the 

 extension of this term to all membeis of the homologous series, 

 on the basis of a common constitution, viz., as aldehydes or ke- 

 tones of the normal polyatomic alcohols of the aliphatic series. 

 As characteristic properties of all these, must follow their be- 

 havior toward polarized light, toward alkaline metallic solutions, 

 and toward phenylhydrazin. 



Such a classification would exclude the bodies of the cellulose 

 group, of which there are many, more or less v\-ell defined. But 

 it is not yet evident that they possess a constitutional relation to 

 the bodies under discussion, and have certainly no claim to be 

 classed with the aldehyde or ketone alcohols because convertible 

 into them. 



As for the disaccharides of the hexoses, to which belong 

 sucrose, lactose, etc., if it be true, as supposed, that they are an- 

 hydrides or ether-like forms of the hexoses, then they are entitled 



