OBSERVED COMPOSITE SPECTRA OF BETA AND

Assembled 10-15-93.

Bet And photometry file: photometry actually used to construct the spectrum.

Name   FWHM  Mag.+/-Unc.  Eff Wvl  Eff Wvl    F-lam
		         (Vega)   (star)     W/cm2/um
      (um)                (um)     (um)
Kn    0.0488 -1.93 0.01   2.208    2.205     2.33E-13
Ln    0.1443 -2.06 0.01   3.782    3.762     3.44E-14
M     0.6677 -1.74 0.02   4.758    4.749     1.04E-14
8.7   1.1576 -1.98 0.01   8.753    8.736     1.21E-15
11.7  1.2008 -2.09 0.01  11.650   11.622     4.28E-16


Spectral fragments and portions of these actually used in observed spectrum
("used" may include combination with other data where overlaps occur).

Fragment   Reference    Total range    Start and stop    Average resolving
			   (um)	       wavelengths (um)        power
NIR           1         1.22- 5.50       1.22- 5.50             50
KAO           2         3.38-10.12       3.38-10.12       50,50,45,70,150
8-13          3         7.59-13.06       7.59-13.06            65,45
LRS           4         7.70-22.60       7.70-16.90             30
LONG          5        15.52-24.09      16.95-23.37             73
VERY-LONG     6         1.25-35.00      12.00-35.00             --

References:
1. Strecker, Erickson, and Witteborn 1979, Ap.J. Suppl., 41, 501
2. FOGS data of Nov. 29, 1983, Dec. 1, 1983, and Dec. 12, 1985 KAO flights
   [Bet And/Alp Tau]; Jan. 18, 1991 FOGS+SIRAS data [Bet And/Alp CMa]; and
   HIFOGS KAO flight of July 21, 1993 [Bet And/Alp Lyr].
3. FOGS Mt Lemmon data of Sept. 17, 1985 [Bet And/Bet Peg] and UKIRT CGS3 data
   of August 29, 1993 [Bet And/Alp CMa] and [Bet And/Alp Tau].
4. LRS raw data extracted from the new Groningen IRAS database and
   recalibrated with "LRSCAL" routine in "GIPSY" package.
5. UKIRT CGS3 20 um ratio of [Bet And/Alp Tau] of Oct. 5, 1990 {M.J. Barlow}.
6. Engelke Fn. used for T=3839K (see Blackwell, Lynas-Gray, and Petford 1991,
   Astron. Astrophys., 245, 567) and ang. diam.=13.219 milliarcsec; we
   rescaled this to 13.71 mas.  This Engelke Function was locked to the
   photometrically scaled combination of 8-13, LRS, and LONG spectra by
   "splicing" and used to replace the observations beyond 12.00 um (bias only
   +/-0.16%).  The uncertainty in EFn. due to an effective temperature error
   of 100K is 2.7% which was input as the error of this fragment because of
   the additional shape constraints available from the LONG spectral fragment.


Information on splices and biases incurred

Process                  Factor determined       +/-Bias%
NIR cf. photometry           0.983                0.90
813 cf. photometry           0.997                0.66
LRS splice to 813            0.941                0.14
KAO joint splice to
 NIR and merged 813/LRS      0.881                0.05
Engelke Fn. splice to
 combined 813/LRS/CGS3-20um  1.075                0.16

 ====================================================
 Notes:
 1. Template spectra are NOT tabulated at equal
    intervals of the wavelength, but rather at the
    wavelengths of the original measurements.
 2. In most cases "total uncertainty" is the error
    term most appropriate to use. It is the standard
    deviation of the spectral irradiance and includes
    the local and global biases.  Local and global
    biases are given as a percent of the irradiance.
    The global bias does not contribute error to
    flux ratios or color measurements, and may, in
    those cases, be removed (in the rss sense) from
    the total error.
 ====================================================