May, 1942
NOTES
1233
TABLE I ADDITION COMPOUNDS OF TETRAHYDROTHIOPYRAN Inorg. salt
Solvent
Solubility
M. p.. OC.
Formula
Metal analyses, % Calcd. Found
Alcohol Alcohol Alc. HCI Alcohol Alc. KBr Alc. K I Ether Alcohol Ether Alcohol
S1. sol. hot aq. alc. 101-105 HgBrc (CHd.3 43.36 43.26 Ins. aq., org. solv. 154.5-157 Cu Cl. (CHz)&S 31.59 31.55 Same 154.5-160 CUCI.(CHZ)~S 31.59 31.50 Same 123-124 CuBr.(CHz)6S 25.88 25.85 Same 121.5-122.5 CuBr. (CH2)& 25.88 25.79 Same 164-165 d. CUI.(CHz)aS 21.73 21.85 SI. sol. alc., ins. eth. 120-122 d. AuCls.(CHz)sS 48.60 48.39 Ins. alc. eth. 179-182 d. AuCls (CHz)sS 58.89 58.66 SI. sol. alc. eth. 140-145 d. AuBw (CH&S 36.58 36.70 a S1. sol. hot alc. 173-179 d. AuBr. (CH&S 51.99 51.83 b Sol. chl., dec. aq. SnCL 149-151.5 SnClr,2(CH&S 25.53 25.57 Chlorof. Sol. chl., dec. aq. 149.5-151 SnBrc2(CH2) SnBrr 18.36 18.45 H~PtC1~-KIC Alcohol Ins. alc. 194.5-196 d. PtIz'2(CHz)sS 29.88 29.98 AcetoneHz0 SI. sol. acet., ins. aq. 146.5-148.5 d. P ~ C ~ Z . ~ ( C H ~27.93 )~S PdClr 28.07 a AuBra.(CH2)sSboiled with alcohol and excess thiopyran. Thiopyran and tin(1V) chloride mixed directly. Excess potassium iodide solution added to chloroplatinic acid in alcohol, precipitate dissolved by heating, and thiopyran added till no further precipitation. HgBrz cuc12 Cu2Clz CuBrz CuzBr2 c!u2Iz HAuClr HAuClr HAuBr4
hydrothiopyran by refluxing with excess sodium sulfide in ethanol. The addition compounds were formed by dissolving the metal salt in ethanol or ether, adding salt or acid if necessary, and adding slight excess of sulfide, sometimes dissolved in the solvent. The data on the complexes are given in Table I.
potassium and ammonium fluogermanates. The single needle-like specimen may or may not have been rubidium fluogermanate : the values of the lattice constants indicate that i t may have been a different crystalline modification of closely related and only slightly more complicated structure than that to be discussed here. In any case we must correct our earlier statement a t least to the DEPARTMENT OF CHEMISTRY UNIVERSITY OF WASHINGTON extent of asserting that the usual and presumably SEATTLE, WASHINGTON RECEIVED FEBRUARY 2, 1942 the stable form a t room temperature of rubidium fluogermanate is fully isomorphous with the corStructures of Complex Fluorides. Rubidium responding ammonium and potassium salts. The X-ray data for Rb2GeFslead to a hexagonal Hexafluogermanate unit cell with a = 5.82, c = 4.79 A., space-group BY W. B. VINCENT AND J. L. HOARD D:d - C%m,containing one stoichiometric moleIn an earlier paper' we have reported the X-ray cule. The atomic coordinates2 are Ge in (a): determination of structure for the isomorphous 000; 2Rb in 2(d): l/s u; tZ with u = crystalline compounds potassium hexafluoger- 0.695; 6 F in 6(i): xxz, etc., with x = 0.144, z = manate and ammonium hexafluogermanate, 0.213. R2GeF6. Upon the basis of X-ray data obtained The methods used in establishing this structure from a single specimen of what we supposed was closely paralleled those previously described. rubidium fluogermanate, we stated a t that time A comparison of calculated with observed rethat this substance crystallizes in a more com- flection amplitudes for over one hundred forms in plex structural type than do the corresponding rubidium fluogermanate led to generally excellent potassium and ammonium salts. This particular agreement. Having presented' the corresponding specimen occurred as a hexagonal prism, almost data in detail for potassium fluogermanate, and needle-like in shape, whereas we have since ob- ammonium hexafluogermanate, it seems untained only hexagonal tablets and plates by re- necessary to reproduce the rather extensive crystallizing rubidium fluogermanate from aque- tables3 for rubidium hexafluogermanate in this ous solution a t room temperature. X-Ray study note. In addition we find that a Fourier projecof a number of these tabular crystals shows that (2) "Internationale Tabellen zur Bestimmung von Kristallstrukthey crystallize in the same structural type as do turen," Gebriider-Borntraeger, Berlin, 1935, Vol. I, p. 258. (1) J. L. Hoard and W. B. Vincent, THISJOURNAL, 61, 2849 (1939)
(3) The amplitude data for rubidium hexafluogermanate are available in the Thesis of W. B. Vincent, "The Structures of Some Fluosilicates and Fluogermanates," Cornel1 University Library. 1940.
COMMUNICATIONS TO THU EDITOR
1234
tion of relative electron density along a using experimental (OM) amplitude data is in satisfactory agreement with the parameter values Riven above. Rubidium Ruogcmianate is an aggregate of Kf and practically regular octahedral GeFsions (for diagrams of the structural type see ref. 1). The lattice constants and parameter values are only very slightly different from those found for ammonium fluogennanate, so that corresponding interatomic separations are virtually identical
VOl.
GI
in the two cas%. The near identity in the effective radii of rubidium and ammonium ions appearing in corresponding compounds has been repeatedly observed excepting in cases where ammonium ion is restricted to a small coijrdination number (usually four) through the formation of strong hydrogen bonds. We may conclude again1 that hydrogen bonding plays a relatively minor role in ammonium Rtiogermanate. D e P A R T M s m OF CIIRMISTRY
CORNELL UNIVERSITY ITHACA.NEW YoRu
RECRIVEU VKBRUARV
zn.
19.r~
C O M M U N I C A T I O N S T O T H E EDITOR ELECTRON MICROSCOPE OBSERVATIONS OF COLLAGEN
Sir: Electron micrographs have been made of collagen fibers from a variety of sources, including rat tail tendon, beef tendons and ligaments, and human skin. Fibers werc obtained either by teasing small bits of tendon in water or by dissolving the material in acetic acid and reprecipitating the fihers by neutralization.
'.Or*
Fig. I.-Electron micrograph of collaacn fihers lrom Ixrf tendon. magnification ?S.IXxI X .
Laboratory [R. S. Bear, THISJOURNAL, 64, 727 (1'342)l have demonstrated the presence in collagen of a fiher-axis periodicity of approximately WO A. This spacing was obtained from all the types of collagen mentioned above and appears to he characteristic of this protein in inlacl tissues. There r e m s little doubt that the periodicities observed in the electron micrographs represent a manifestation of the X-ray diffraction periodicity i n intart tendon and that the phenomenon is a consequence of the structure and arrangement of the collagen molecules in the fibers. The range of spacings observed in the electron micrographs is dnubtless due to the special conditions required fnr the preparation of the material, chief among which arc the isolation and vacuum drying of individual fibers. I t is reasonablc to expect individual fibers to behave different y when isolated than when present in compact bundles as in normal tendon where lateral restraints, possibly by enclosing membranes and cement substance, restrict their behavior mechanically. This interpretation is being tested by an X-ray diflraction study of teased fibers similar to those observed with the electron microscope. In addition, the effect of various physical and chemical conditions on the appearance of the fibers in the electron micrographs is being further investigated in an effort to get more information concerning the molecular architecture of collagen. Dnm. OP B~OLOOY AND PUBLIC H ~ A L T I I CECILE. HALL
Under appropriate conditions the fibers appear characteristically cross-striated, the relatively opaque and transparent bands extending uniformly across the fiber (see Fig. 1). The average distance between like bands can be measured with an accuracy of about 3%. The interband distance is independent of fiber width and varies considerably from one fiber to another; the extremes thus far measured are '3013 and 522 A,. though the range shown in the fibers of any single MASSACHUSETTS I ~ s r m ~ OF t nT R C I I . MARIRA. JAKUS preparation is more restricted. CAMBRll%E. M A S S . mAh'CIS 0.S c l I M l T T Recent X-ray diffraction investigations in this RECE1vsD MARCN IT. 1942
