weissler_1952#
G. L. Weissler and Po Lee, “Absorption coefficients of oxygen in the vacuum ultraviolet”, J. Opt. Soc. Am. 42, 200 (1952)
Source record · Catalogue · Review register
Source kind: measurement
Recorded technique: Photographic photometry with a grazing-incidence vacuum spectrograph filled with O2 at 0.5-100 micron (Hg) and a line-emission light source (lines of O, N, C, Si, H, He ions, some in second and third order); light path 79.7 cm at 550 A to 103.3 cm at 1300 A. Each k is the slope of ln I against p from at least three, often five, pressures (Beer’s law tested) (Eqs. 1, 1a, pp. 200-201). Wavelengths are those of the source lines (vacuum ultraviolet: vacuum wavelengths; the printed wave numbers are 1e8/lambda). Resolution: that of the spectrograph with a line source; not stated.
Materials in structured entries: O2
Dataset quantities: absorption_cross_section, oscillator_strength
Contents: 4 datasets; 1 formula.
Curation notes#
Journal pages 200-203 = PDF pages 1-4. O2 only; 76 rows between 303.779 and 1306.038 A (the paper’s ‘76 wavelengths’ counts the OI 990.797 line printed twice).
Quantity: k (cm^-1) is defined by I = I0 exp(-k x), ‘x is the thickness of the absorbing gas layer at N.T.P.’, reduced with the ideal-gas law x = T0 L p / (p0 T), T0 = 0 C, p0 = atmospheric pressure (Eq. 1, pp. 200-201): a natural-log absorption coefficient per unit length at 0 C and 1 atm. The paper gives no conversion to cross sections. The datasets store the printed k with scale = 1e18 / n0 (and x 10^2 for Table I’s column factor), n0 = 101325 Pa / (k_B 273.15 K) = 2.6867801e19 cm^-3 (the ideal-gas Loschmidt number from the exact SI constants, matching the paper’s ideal-gas reduction), giving the photoabsorption cross section per molecule in Mb. The vocabulary has no photoabsorption cross section: ‘photoionization_cross_section’ is used, as for the noble gases, although below the first ionisation limit (about 1027 A) the absorption is into bands and dissociation continua, not ionisation (format need reported).
Accuracy (p. 202): ‘The error in the coefficients presented here is estimated to be about ten percent for k-values greater than 100 cm^-1 and fifteen percent for lower values’, with the note ‘The experiment was not designed to measure k-values less than 20 cm^-1 accurately’; slit-width errors ‘within the stated limits’. ‘it is felt that the accuracy of k-values which fall within molecular absorption bands is less than those within a continuum. Furthermore, their physical significance is not clear’ (p. 201): those rows carry a band location (B.H., B.M., B.T.) in the printed column; they are not band averages but values at single source lines, partly inside unresolved bands.
Text values (pp. 200-203, not extracted as data): at 1300 A ‘k less than 10 cm^-1’; between 1300 and 1000 A about 10 cm^-1 outside bands, ‘a weak continuous absorption ... of the order of k = 50 cm^-1 to k = 100 cm^-1’ (p. 202); between 1000 and 740 A the lowest value outside a band 110 cm^-1, in the bands 600 cm^-1 or more; below 740 A k = 500-1000 cm^-1 ‘with a broad maximum of k = 700 cm^-1 between 400A and 600A and still being strong at 303A, namely k = 550 cm^-1’ (p. 202). The abstract says 530 cm^-1 at 303A, as Table I (5.3 x 10^2 at 303.779 A): the text’s 550 is a misprint of the paper (no dataset holds it). The dotted curve of Fig. 1 (figure only) joins roughly the lowest k values between 1000 and 300 A. Reproduced (p. 200): Ladenburg and Van Voorhis’s Schumann continuum maximum 490 cm^-1 at about 1450 A.
Comparison: Watanabe et al. (1953, watanabe_1953) found the k of these authors in 1060-1170 A ‘in general, considerably larger’ (attributed to resolution) and the weak continuum near 1105 A of order 1 cm^-1, not 50-100 cm^-1; Metzger and Cook (1964, metzger_1964) measured windows of k = 0.20-0.62 cm^-1 at 1108-1216 A, where this table gives 43-120 cm^-1 at nearby lines. Above 1027 A these values are not reliable between bands.
Wave-number and wavelength columns disagree (1e8/lambda against the printed wave number) beyond their precision in 14 rows; the wavelength (the source line) is used and only row 54 is corrected, where the paper itself gives the evidence (erratum). Rows (printed wave number; wavelength implied by it): 4 (80431.1; 1243.30), 8 (83364.8; 1199.55, italic digit), 15 (92201.7; 1084.58), 17 (96497.1; 1036.300), 18 (97942.8 for 97492.8, a transposition), 32 (110667; 903.611), 33 (113815; 878.62, against 886.618 printed), 38 (120060; 832.917), 48 (142284; 702.820), 52 (148600; 672.948), 54 (151450; 660.284: erratum), 63 (172177; 580.797), 68 (185499; 539.086), 71 (188900, italic 00; 529.38), 74 (235236 for 232536, a transposition). In bands and windows a wrong wavelength would matter; in the continuum below 740 A it does not.
Table II’s percent differences are not all reproduced by its columns: 640 and 670 differ by 4.6 %, printed 7.8 (the others agree to 0.3 %); not extracted.
Formula: Ladenburg and Wolfsohn’s three-term dispersion of O2, quoted p. 200 (see formulas); the paper’s integral of its continuum (p. 203) ‘a mean oscillator strength of approximately f = 6’, against Ladenburg and Wolfsohn’s f3 = 5.9.
Recorded bibliographic identity status: VALIDATED_WITH_WARNINGS. This concerns identification of the source; it does not establish the scientific correctness of its contents.
Datasets#
weissler_1952/table_1#
{
"id": "table_1",
"material": "O2",
"file": "O2/weissler_1952__table_1.csv",
"table": "tbl_56bc106d3886, tbl_3c7b2b7dceeb",
"quantity": "absorption_cross_section",
"unit": "Mb",
"scale": 3.721927208,
"abscissa_unit": "angstrom",
"wavelength_medium": "vacuum",
"use": "primary",
"excluded_rows": {
"1": "k = 7 cm^-1, below the 20 cm^-1 the experiment 'was not designed to measure ... accurately' (p. 202, Table note)",
"24": "the paper prints the line OI 990.797 (100929 cm^-1, 1.7 x 10^2) twice, as rows 21 and 24; the repeat is dropped"
},
"uncertainty": {
"relative": 0.1,
"level": "not stated",
"relative_rows": {
"2": 0.15,
"3": 0.15,
"5": 0.15,
"7": 0.15,
"8": 0.15,
"11": 0.15,
"12": 0.15,
"13": 0.15,
"15": 0.15,
"18": 0.15
}
}
}
Table I, rows in print order 1-76: k (cm^-1, natural log, layer reduced to 0 C and 1 atm) at the source-line wavelengths, printed as mantissa x 10^2; scale = 1e20 / n0 (n0 = 2.6867801e19 cm^-3) turns the mantissa into the photoabsorption cross section in Mb (see the source notes). 10 % for k > 100 cm^-1, 15 % for k <= 100 cm^-1 (rows with mantissa <= 1.0). Rows with a band location (printed column: B.H., B.M., B.T., reference 6 Price and Collins or 8 Tanaka and Takamine) lie inside molecular bands; the paper doubts their accuracy and meaning. Values above about 1000 A disagree strongly with the photoelectric data of watanabe_1953 and metzger_1964 (source notes). Row 54 prints 660.380 A for 660.280 (erratum).
weissler_1952/first_order#
{
"id": "first_order",
"material": "O2",
"file": "O2/weissler_1952__first_order.csv",
"table": "tbl_3deee1da1d4e",
"quantity": "absorption_cross_section",
"unit": "Mb",
"scale": 0.03721927208,
"abscissa_unit": "angstrom",
"wavelength_medium": "vacuum",
"use": "check",
"uncertainty": {
"relative": 0.1,
"level": "not stated"
}
}
Table II, ‘First order’ (plate LE 25): k (cm^-1) from the first-order lines, compared with the second-order ones; scale = 1e18 / n0 to Mb. A consistency test of the photometry in two grating orders, not fitted. The OII row prints 644.188 A for the 644.148 of Table I (erratum).
weissler_1952/second_order#
{
"id": "second_order",
"material": "O2",
"file": "O2/weissler_1952__second_order.csv",
"table": "tbl_3deee1da1d4e",
"quantity": "absorption_cross_section",
"unit": "Mb",
"scale": 0.03721927208,
"abscissa_unit": "angstrom",
"wavelength_medium": "vacuum",
"use": "check",
"uncertainty": {
"relative": 0.1,
"level": "not stated"
}
}
Table II, ‘Second order’: three of the four values equal Table I’s second-order rows (750, 720 and 780 cm^-1 at 644.825, 645.167 and 660.280 A); the OII line gives 670 here and 690 in Table I. A duplicate of Table I’s measurement: a check, not fitted. The OII row prints 644.188 A for 644.148 (erratum).
weissler_1952/continuum_f#
{
"id": "continuum_f",
"material": "O2",
"table": "text",
"quantity": "oscillator_strength",
"abscissa": "label",
"use": "check"
}
1 inline points in the source record.
p. 203: ‘When integrating graphically the dotted curve in Fig. 1 representing continuous absorption, a mean oscillator strength of approximately f=6 was found’, agreeing with Ladenburg and Wolfsohn’s f3 = 5.9. The dotted curve joins ‘roughly the lowest measured k-values between 1000A and 300A’ and is ‘probably a reasonable approximation to the continuous absorption’ below 740 A (p. 202); the range of the integral is not stated. Index abscissa: an integrated strength, not a line. Approximate; a check.
Formulas#
weissler_1952/ladenburg_wolfsohn#
{
"id": "ladenburg_wolfsohn",
"material": "O2",
"equation": "text, p. 200 (reproduced)",
"printed": "represented by a three term equation containing three characteristic wavelengths, lambda_1 = 1898.6A, lambda_2 = 1467.9A, lambda_3 = 544.36A, and three corresponding oscillator strengths, f_1 = 4.04 x 10^-5, f_2 = 0.202, and f_3 = 5.93",
"variable": "lambda: wavelength in A (the equation itself is not printed)",
"state": "not stated",
"density": "unknown",
"range": "the interferometric dispersion measured between 5900 A and 1920 A (p. 200)",
"coefficients": {
"lambda_1": "1898.6",
"lambda_2": "1467.9",
"lambda_3": "544.36",
"f_1": "4.04e-5",
"f_2": "0.202",
"f_3": "5.93"
},
"reproduced_from": [
"R. Ladenburg and G. Wolfsohn, Z. Physik 79, 42 (1932)"
],
"check": "Read as a per-molecule oscillator model alpha = sum f_i / (E_i^2 - omega^2) (atomic units, E_i = hc/lambda_i), it gives the static polarizability 10.562 au and 10.774 au at 589 nm, 11.510 au at 300 nm, 13.694 au at 200 nm: 0.24-0.45 % below the stored fit/O2 (10.587, 10.805, 11.550, 13.755 au) (computed 2026-09-25)."
}
Quoted without the equation; the f_i are classical oscillator strengths per molecule. Watanabe et al. (1953) quote lambda_2 = 1468A from the same work and Koch’s (1912) lambda_2 = 1421A. The original is not in the library (missing-K).
Corrections#
[
{
"dataset": "table_1",
"row": "54",
"column": "x",
"printed": "660.380",
"corrected": "660.280",
"reason": "The second-order NII line is printed 660.380 A in Table I; the same line is 660.280 A in Table II, and Table I's own wave number, 151450 cm^-1, is 1e8/660.280 (660.380 would be 151428).",
"evidence": "Table I (page 3 continuation, image tbl_3c7b2b7dceeb.jpg) and Table II (tbl_3deee1da1d4e.jpg), checked 2026-09-25; Table I's k (7.8 x 10^2) equals Table II's second-order 780 cm^-1 at 660.280."
},
{
"dataset": "first_order",
"row": "644.188",
"column": "x",
"printed": "644.188",
"corrected": "644.148",
"reason": "Table II prints the OII line as 644.188 A; Table I has no such line but OII (second order) 644.148 A with the wave number 155244 cm^-1 = 1e8/644.148 (644.188 would be 155234). The other three Table II lines are Table I lines.",
"evidence": "Tables I and II, images checked 2026-09-25. Moderately certain: Table II's second-order k (670) differs from Table I's (690) for this line, while it equals Table I's for the other three."
},
{
"dataset": "second_order",
"row": "644.188",
"column": "x",
"printed": "644.188",
"corrected": "644.148",
"reason": "As for first_order: the OII line of Table I is 644.148 A (wave number 155244 cm^-1).",
"evidence": "Tables I and II, images checked 2026-09-25."
}
]
Formula corrections, where recorded, appear with their formulas above. Curation notes may also discuss corrections or unresolved interpretation. Consult the review register for the checks and decisions.