_ THE PHILIPPINE JOURNAL OF SCIENCE VOLUME 19 JULY TO DECEMBER, 1921 WITH 72 PLATES AND 27 TEXT FIGURES senor" eantiats, A - nud 7 1922 : \ / EDITORIAL BOARD ELMER D. MERRILL, M.S., Editor R. C. McGrecor, A.B., Associate Editor ALBERT H. WELLS, A.B.; GRANVILLE A. PERKINS, PH.D.; A. P. West, PH.D. T. Dar JUAN, A.B., PHAR.D.; F. AGcaorti, A.B.; A. S. ArGUELLES, B.S. ALBERT E. W. KING Chemistry WARREN D. SMITH, PH.D.; Roy E. DicKERSON, PH.D. Geology H. W. WADE, M.D.; Otto ScHést, M.D. F. G. HAuGHWouT; STANTON YouNGBERG, D.V.M. Experimental Medicine LipoR10 GOMEZ, M.D., PH.D.; F. CALpERON, B.A., L.M. * VICENTE DE JESus, M.D. Clinical Medicine W. H. Brown, Pu.D.; C. F. Baker, M.A.; H. ATHERTON LEE, M.S. O. A. REINKING, B.S.A., M.S.; L. M. GueRRERO, PHaR.D. Botany ALBERT C. HERRE, PH.D.; C. F. Baker, M.A.; S. F. Licut, M.A. | C. S. Banks, M.A.; L. D. WHarton, M.A.; W. SCHULTZE Zoblogy H. O. Breyer, M.A.; Otto JOHNS SCHEERER, M.A. Anthropology ANNA B. BANYEA, Copy Editor ii CONTENTS No, 1, July, 1921 {Issued October 31, 1921.] PERKINS, GRANVILLE A. The expression of the octet theory of va- lence::in: structural formulas 2. bic e edntn One plate. LicHt, S. F. Notes on Philippine termites, II..............-..-..-------------- Six plates and three text figures. Starr, O. A new species of Vincentia from the Philippines................---- Weis, A. H. The preparation of tikitiki extract for the treatment yl Pitre RN acacia kde pean poets ronleneninins ceases Rouwer, 8. A. The Philippine wasps of the subfamilies Scoliine and Elidine REINKING, OTTo A. Higher Basidiomycetes from the Philippines and their hosts, V............----::-0:-sssseccscscsceesesnesenenscscetnenenereennesscnesnenanenecees McLean, F, T. The permeability of citrus leaves to water................ One text figure. REVIEWS No. 2, August, 1921 [Issued December 9, 1921.) Lee, H. ATHERTON. Citrus-canker control: a progress report of experiments ¢ Two plates. ScHEERER, OTTo. Kalinga texts from the Balbalasang-Ginaang group One text figure. WILEMAN, A. E. Notes on Japanese Lepidoptera and their larvae: Part VI sels Two colored plates. g SMITH, WARREN D. Ancient cave dwellers of Batwaan, Masbate, Philippine Islands. Five plates and two text figures. WHARTON, LAWRENCE D. Opisthorchis wardi, a new species of liver fluke from the cat in the Philippine Islands One plate. Licut, S. F. Notes on Philippine Alcyonaria, Part VI: New Philip- pine Pennatularia (sea pens) of the genus Lituaria.........-- . iil . Page. 115 127 129 175 209 233 243 247 iv Contents No. 3, September, 1921 [Issued December 19, 1921.] MAXWELL, J. PRESTON. Filariasis in China..............................-....--. Twenty-five plates and four text figures, MERRILL, ELMER D. A review of the new species of plants proposed DYN... -Barame in: his Flora liidiea...30...000 Ss No. 4, October, 1921 [Issued January 16, 1922.] REINKING, O. A., and Grorr, G. W. The Kao Pan seedless Siamese pommelo” and ite cultwve. se Sixteen plates and one text figure. GEBIEN, HANS. Philippine Tenebrionidew, II................................. Two plates. BANKS, CHARLES S. A Philippine nemestrinid Ceere ys... One plate. No. 5, November, 1921 [Issued February 15, 1922.] HELLER, K.M. New Philippine Coleoptera... Three plates. No, 6, December, 1921 [Issued March 20, 1922.] LANTIN, PEDRO T. Various methods of serum application in bacillary dysentery io os Six text figures. PERKINS, GRANVILLE A. Unsymmetrical addition to the double bond, I: A theory of the reaction mechanism of the direct union........ Six text figures. GROVES, JAMES. Charophyta from Annam and Guam ett ete tte Teer erry RoHWER, S. A. The Philippine wasps of the subfamily Sphecine.... MERRILL, ELMER D. Two new species of plants from Hainan... ee Cote, Howarp IrRvinG. The dissociation of hexaphenylethane from the viewpoint of the octet theory of valence. McGrecor, RicHARD C. New or noteworthy Philippine birds, IV.... Four plates and three text figures. BECKER, WILHELM. Die Violen der Philippinen OxKAmoTo, Kikuo. Secondary sexual characters in the loach Misgur- nus anguillicaudatus Cantor. One plate. Saree eee rene tema eweaeeeneee PERKINS, GRANVILLE A. The structure of chlorine dioxide and relat- Oe vornpounds oo 389 439 523 629 723 729 Contents Vv Page WELLES, CoLIN G. Cercospora leaf spot of coffee... 741 One plate. WELLES, COLIN G. Cercospora leaf spot of Averrhoa carambola... 747 Two plates. PURVEEW iin... ee ee a a a 753 RORMRAT A och tec ee ee ea a es 755 INDEX. oo ae a ee eas 757 THE PHILIPPINE JOURNAL OF SCIENCE VoL. 19 JULY, 1921 No. 1 ~ THE EXPRESSION OF THE OCTET THEORY OF VALENCE IN STRUCTURAL FORMULAS? By GRANVILLE A. PERKINS Chemist, Bureau of Science, Manila ONE PLATE The science of organic chemistry, as we know it to-day, may be said to owe its very existence to the idea of structural formu- las developed by Kekulé, Frankland, and Couper, sixty years ago. In the last two decades, however, the development of both organic and inorganic chemistry has been greatly retarded by the fact that Kekulé’s simple “affinity units” do not represent with sufficient accuracy the actual forces which bind atoms together. Recent attempts of physicists to apply the new knowledge of electrons to the fundamental problem of chemistry, namely, the nature of chemical affinity, met with only partial success, until Langmuir? finally showed that certain recent ideas of atomic structure, notably those of Lewis,* can be used to form a re- markably successful working hypothesis in both organic and in- organic chemistry. The conception of electron shells and shared electrons as presented by him gives one a definite picture which is undoubtedly very close to the actual nature of the union between atoms. Langmuir’s “octet theory of valence” is so simple and exact that it will be found of great value to chemists, from the over- curious student who asks his professor what makes the atoms stick together to the investigator who wishes to predict the "Received for publication, February 28, 1921. * Langmuir, I., Journ. Am. Chem. Soc. 41 (1919) 868, 1543; 42 (1920) 274. * Lewis, G. N., Journ. Am. Chem. Soc. 38 (1916) 762. 180365 1 2 | ‘The Philippine Journal of Science 1921 results of a reaction which has never been performed. The method used by Langmuir in applying the theory, however, ap- pears to the writer to be somewhat cumbersome for general use, and to obscure, by its indirectness, some of the value of the theory. The purpose of the present article is to present a sys- tem for writing structural formulas which will be as simple and direct as possible, and at the same time represent the mole- cules as accurately as possible in terms of modern atomic theory. It is hoped not only to furnish by this means a simple method for practical application of the theory in its present stage of advancement but: even to develop certain valence relationships which have not hitherto been clearly expressed. While it is assumed that anyone interested in the subject is already familiar with recent developments of the octet theory in the hands of Langmuir and others, for the sake of clarity the subject will be briefly reviewed. THE STRUCTURE OF ATOMS NUCLEUS ‘ In the light of recent physical evidence the essential portion of any atom is a minute nucleus composed of positive units of ‘electricity (sometimes called positive electrons) and a smaller number of negative units (sometimes called negative electrons, but usually simply electrons) very closely packed together and bound by the most powerful forces known. The positive units are all identical, each having a mass of approximately 1, ex- pressed in atomic weight units. Similarly the (negative) elec- trons are all identical, but have negligible mass. Each electron in the nucleus neutralizes one positive unit, so the total outside electrical effect, called the nuclear charge, is measured by the number of positive units less the number of electrons (negative units). This difference is called the atomic number of the nu- cleus. NEUTRAL ATOM Except under very unusual circumstances, such as when trav- eling with extreme velocity, a nucleus is never found alone, be- cause it normally attracts as many electrons as its atomic num- ber, forming an electrically neutral atom. These electrons do not enter the nucleus but arrange themselves in nearly spherical concentric shells, which vary in number from one to seven ac- ‘Harkins, W. D., Journ. Am. Chem. Soc. 42 (1920) 1956; Rutherford, E. E., Proc. Roy. Soc. London 97A (1920) 374-401. 19, 1 Perkins: The Octet Theory of Valence 3 cording to the number of electrons. The shells are filled in the following order: First shell, 2 electrons; second, 8; third, 8; fourth, 18; fifth, 18; sixth, 32; seventh, (32?). (See Table 1). The electrons either rotate, perhaps in ring form, or revolve in small* orbits, so that they are powerful electro-magnets. ‘(See Plate 1.) TABLE 1.—The atoms arranged according to their electron. shells. 3 Hydro-| First |Second/ First |Second| Rare | Uran- gen | short | short | long long | earth | ium period. | period. | period. | period. | period. | period. | period. | Bae eh Elec- Vacen-| tronsin} cies in Number of shells______ : 1 2 3 4 5 6 7 | shell. | shell. | Electrons in kernel-__- 0 2 10 18 386; 64 86 Completed shell _______ 2 8 s 18 18 82} (82?) (* O) | (He2) |(Ne10) | (A 18) |(Kr 36) |(Xe 54) |(Nt 86) | ena (H1)| Li3|.Nall/ K19| Rb37| Cs55| * 87 pe oe eers Be4 | Mg12; Ca20| Sr88| Ba5dé/} Rags le Hier B5| Ali3/} Se21| Yt39! La57|} Ac89 __ Biaath aoe Ti22) Zr40|/ Ce68/ Th90 RA ee eA V 23) Cbh41| Pr5d9 |Ux291 Biteociccns Cr 24 | Mo42| Nd60| Ug2 3 ee Mn 25 * 43 pie > Ue Hoenn Spa T loveckgae Ba GS to. Bs or aE 6G G8 teas 2c; Le aaa 90) Gd 6th occ PU ieee ees PUGS (oo jigs a Ho 66 1.2 bi ee 2 Bae DY OFA vaendence ee D 7, o,f eee geno Be becascoe yin By Recent Ses 1B feel Tme to pean AR i303 po eD Gicdacstiace th gy beer aE ta TE fe 10 feces coal Tee eke ID) scudese | We TE tee io ry Peer EES Ammar pe eens Fe 26} Ru 44| Os 76 |__..-__- (Se joc. Ceo? | Rh 45) ir ?7 |... (a8 Ni2s|) Pd i6| Pt 78 |._---.-- (10) 345 Cu 29 | Ag 47 | Au 79 |________ CS Tic Zn 30 | Cd 48 | Hg 80 |_______- cP bide maaan Ga31| In49/| T181 |_-..-... (18)27 |..i....-} C6} Sil¢| Ges2| Sn60| Pb& |__-. 22]. ea | NY| P16 | Ass | Obit | Sige 3) O8 S16/| Se84| Te52 RaF 84 abe athe 2 Hi F9| Cli7} Br35 153 bode Se peels emsenees a a a 1 He2 |Nel0} Al18 | Kr86/| Xe54 | Nt86 |___-____|______.. 0 * Not yet discovered. * Concerning the size of these rings or orbits, there is much difference of opinion among physicists. They may be much smaller, relative to the shell, than represented in Plate 1, but larger orbits than these seem to be excluded by chemical evidence. 4 The Philippine Journal of Science 1921 KERNEL ° ¢ The completed inner shells of an atom (that is, all the shells except the outermost) together with its nucleus, constitute the kernel. This is bound together by such strong electrostatic and magnetic forces that it is never disrupted in ordinary chem- ical reactions, but acts as an unchangeable unit except under extremely penetrating forces, such as those of X-rays. SHELL For brevity the outer shell is called simply the shell. Unless this is complete, as in the helium group, the neutral atom exerts forces, both magnetic and electrostatic, effective at a considerable distance. It is upon these forces that all chemical action depends. This is, roughly, the theory of atomic structure which seems to meet with general acceptance among chemists at the present time. CHEMICAL UNION Two types of chemical union have been distinguished; namely, primary valence unions and secondary valence unions. The present paper deals chiefly with the primary type, which involves the main uniting forces of atoms. This has been subdivided into two kinds, which may be called salt-forming unions and direct unions. Salt-forming unions are caused by the fact that an atom with a nearly complete shell has a strong tendency toward completing its shell, and is able to appropriate from another atom with a less complete shell one or more electrons. Both atoms thus become charged and attract each other. This type of union has often been represented as follows: Direct unions have been a much more puzzling problem, but Lewis’ has advanced the explanation, later developed by Lang- muir, that they are brought about by the sharing of one or more pairs of electrons by two atomic shells. For such a union to take place it is necessary that both atoms have a tendency to complete their shells. Since, in such a case, neither can detach an electron from the other, they compromise by sharing 2, 4, or 6 (but not se * Lewis, G. N., loc. cit.; cf. Bohr. N., Phil. Mag. 26 (1913) 1476; Parson, A. L., Smithsonian Misc. Coll. 65, No. 11; Milliken, R. A., Science 45 (1917) 321; Silberstein, L., Phil. Mag. 39 (1920) 46. "Loc. cit.; cf. Parson’s “group of fourteen,” op. cit., p. 29. 19, 1 Perkins: The Octet Theory of Valence 5 3, or 5)* electrons, which then do duty in both shells. Thus two chlorine atoms, having incomplete shells of 7 each, can form two complete octets in which one pair of electrons is shared. This kind of union, which will be called in this paper a direct union, was represented graphically by Lewis as follows: : Cl: Cl: Langmuir has simplified the formulas, using a single line to represent a shared pair: CI—Cl, HO—CI—O, O—N—O—N—0O, —=N=0, etc. He has shown that a large number of hitherto perplexing structures can be readily expressed in this way. VALENCE Chemists have always tried to make generalizations concern- ing the valence (that is, the quantitative combining power) of atoms. An inspection of the Langmuir formulas above tells one nothing of the true valence relations of the atoms. He calls the number of lines attached to any atom (that is, the number of pairs which it shares in that compound) its covalence, and has developed some valuable generalizations regarding this variable property. What have been considered the true valence relations, however, he consigns to the equation: ° e == 8n — 2p. It seems to the writer that there are two different kinds of direct union not hitherto distinguished. By recognition of this difference it is possible to construct formulas of considerable graphic value, at the same time dispensing with the equation. The proposed system is based on positive and negative valence, the maximum values of which are clearly expressed by Langmuir ® as follows: ‘Now the maximum positive valence is a definite conception—it repre- sents the total number of available electrons in the shell * * * Qn the *In the case of benzene and similar compounds, however, the writer believes that 3 electrons are shared. * This equation expresses the fact that in a molecule the total number of outer shell electrons (e) exist in completed octets, which requires 8 electrons per octet (8x) less 2 for every pair shared (—2p). The inter- pretation of the letters is somewhat modified for the case of hydrogen, where the completed shell has 2 electrons, and the heavy atoms, whose completed shells have 18 or 32 instead of 8 electrons. * Langmuir, I., Journ. Am. Chem. Soc. 41 (1919) 926. 6 The Philippine Journal of Science 1921 other hand, the maximum negative valence represents the number of electrons which the atom must take up to reach a stable form like that of the inert gases. Langmuir further brings out the facts, at least by inference, that the actual negative valence in any compound, if exhibited at all, is almost always the same as the maximum, but the actual positive valence is often less than the maximum, giving two or more classes of compounds of the same metal, such as the cu- prous and cupric compounds. It seems important to the writer to express graphically, when writing structural formulas, the actual positive or negative va- lence exhibited by each atom. PROPOSED SYSTEM FOR WRITING STRUCTURAL FORMULAS 1. Valence.—Represent the maximum electronegative valence of an atom in the ordinary way: | an fF —D—"—N— —C— —s— rN, Each line represents the organic chemist’s “unsatisfied bond,” the physical interpretation of which is vacancy for one more electron in the shell. The nitrogen atom, for example, needs three electrons to complete a shell of eight. The number of “unsatisfied bonds” for each electronegative atom can readily be found by consulting Table 1. Electropositive valence is on no account to be represented by “unsatisfied bonds” as it never causes direct union between atoms. It may be represented as follows: Na!, Cat, AW, 2. The salt-forming union.—It is evident that an atom may fill a vacancy in its shell (satisfy a bond) by simply acquiring an electron, thus becoming a negative ion.*t It may even create more bonds by the reverse process: F— becomes F , —O— be- oe comes O , H— becomes eee ae — N — becomes — Reads: | Negative ions having no unsatisfied bonds form stable saltlike compounds with metallic ions: It is to be noted that while the formation of ions is due to the shell-completing forces of one of the atoms involved, the union " For the lack of a better word, this term seems to be quite generally used for a charged atom even though it may not be mobile. 19,1 Perkins: The Octet Theory of Valence 7 between the ions is due to electrostatic attraction. This is the salt-forming unioh, and will be represented by a dotted line be- tween the ions. 3. The normal direct union.—When two atoms are held to- gether due to the fact that the shell-completing forces of both atoms act on a pair of electrons which is shared between them, the union may be said to be direct. Such a union is to be repre- sented by the usual line for a “satisfied bond” if it is normal; that is, if one electron of the shared pair has been supplied by each atom. In this case it is evident that each shell involved has filled one electron vacancy by the process of sharing and has thus “satisfied” one “bond.” - Examples: H—H, H—O—H. 4. The borrowing direct union.—A direct union in which one atom supplies both electrons of the shared pair may be called a borrowing union. In this case the borrowing atom fills two vacancies in its shell (that is, satisfies two bonds), and the lending atom neither gains nor loses electrons. A convenient way of representing such a union, whereby —-A— satisfies two bonds, and B none, is A>oB. Thesign »has here neither its mathematical nor its astronomical significance, but has consider- able graphic value in representing that two “valence bonds” of A and none of B, are satisfied. ' Cl\. Examples: H — C1O-~ Cl Gi THE CARBON GROUP | | ul ul | oa a hae (Ge ), Sn Fes ee (ap (—Pb—). Vacancies in shell — 4. There is unmistakable evidence that there are definitely four vacancies in the shell of each of these atoms. Sn" and Pb" are predominantly electropositive; but, as the relation of these atoms to the “inert group” has been established by physical evi- dence, each containing 4 less electrons than the corresponding inert atom, we have good reason to believe that this fact plays a large part in determining their chemical properties. This idea : | is substantiated by the behavior of — es —and — Pb— in | organic compounds. The ordinary carbon-hydrogen-oxygen compounds are ex- pressed in the proposed system exactly in the same manner as is customary among organic chemists. The peculiar compound CO, however, has never been successfully explained except by Lang- muir’s hypothesis that the two kernels share 5 pairs of electrons, 1 pair being held rather closely by the kernels and the other 4 pairs in an octet external to both kernels. In this case the va- lence of each atom is increased by 2, because a stable arrange- ment of 10, instead of 8, shell electrons is formed. EXAMPLES cago U=C=6°% 0-0-2 o H Cl ta | ; | Sn === @|- H—Ge—H Cl— Sn—Cl | H Cl 19, 1 Perkins: The Octet Theory of Valence 17 18 " a H—C—C—O-H ==> | fi H H THE NITROGEN GROUP | | bm | Boy if pp eeiaeg eitow ah ig ye cay i Vacancies in shell =3. The system of structural formulas used by organic chemists has never been applied successfully to compounds containing “pentavalent” nitrogen. According to the proposed system ni- trogen never shows in these compounds a valence above 3, al- though, as pointed out by Langmuir, the covalence is 4. The three cases where nitrogen has a peculiar valence of 5 are (=) = (22) N =N, N = O, and — C = N and are explained by Langmuir as being “‘isosteric” with C=0. The cases where nitrogen has been supposed to show a high valence are explained either by the borrowing union or by the formation of a positive ion. The formation of ammonium salts is strictly analogous to the reaction between H—Cl and H—O—H, described in connection with hydrogen. The ammo- nium ion, however, is much more stable than OH, Poe) yay | 4 He NN + ce. Cl .- | H In the case of nitrogen oxides and oxy-acids N has such a tendency to a covalence of 4, that it lends electrons, especially to —O—, as: | O 5 pees 6, NxoN=O 8 | I O=N-—-N=0 Ow~mN-NxO O O H—O-N~N—-O-NcO O 18 The Philippine Journal of Science 1921 It would not be expected, however, that — N — could borrow from a strongly negative atom as — O —, forming such a com- poundas N sO N. In a few compounds the heavier atoms of this group show elec- tropositive tendencies. The salts of Bi™ are good examples of how the weak forces toward completion of a shell may be entirely overcome by some stronger shell, and electrons lost, rather than gained, by the weakly negative atom. It has often been considered that in the formation of pentahalides the atoms of this group show a positive valence of 5. A comparison with vanadium, which of all the positive atoms has the greatest simi- larity to the trivalent negative atoms, is of interest. The most evident difference is that the halides and oxyhalides of this group indicate valences of 3 and 5, but never 4. Vanadium, as would be expected from the consideration of electrostatic forces, shows also a valence of 4. © It is evident then, that the trihalides of this group have a stability which does not permit of the addition of one more halo- gen atom. When two are added they probably form a compound strictly analogous in structure to NH,Cl in most, if not all, cases..7_ Thus, PCl, is tetrachlorophosphonium chloride: Ol Cl a eg PSs oF ok Cl Cl THE OXYGEN GROUP —O— —S-—, —Se— —Te~—, Ra F. Vacancies in shell 2. In most compounds containing borrowing unions it is found that either —O—, or —S— ig the borrowing atom. This is not surprising, as of all the atoms capable of acting in this way these two are the most electronegative. It is the borrowing union which explains the fact that the number of oxygen atoms which can attach themselves to a negative atom depends more on the size of that atom than on its valence. Thus we have ClOg,,. 80, *, PO =, S- t NOs , and CO; ~. The sulphonium and oxonium salts, so familiar to organic chemists, are assigned formulas analogous to those of ammo- nium salts. “ef. Langmuir, ibid, p. 919. 19, 1 Perkins: The Octet Theory of Valence 19 In comparing the halides of this group with those of Cr, Mo, W, and U, we find even stronger indications of definite shell- completing action than were noted in the preceding group. There seems to be no other explanation for the avoidance by S, Se, and Te of the apparent valences 3 and 5 except that they retain always their negative valence of 2, attracting halogen atoms beyond this amount only in pairs, and in the same way that NH, combines with HCl. EXAMPLES 0x8=0 ye 05 ee Di 3 O»Pad Hs O 3 re) CH, F F os y VY é Hos ae | ee ee ee Pash, | CH; F F C] THE HALOGENS F—, Cl—-, Br—-, I~. Vacancies in shell =1. As these are typical negative atoms, there is no difficulty in representing their compounds. F +- F— Ff F— Br —F ae eel oe ie F F Cl 0 O O + 8 8 | tO) H—O—I