A COLD-FINGER DEWAR FLASK WAS USED TO MAINTAIN SAMPLES NEAR LIQUID NITROGEN TEMPERATURE (77 DEGREES KELVIN) DURING LUMINESCENCE ANALYSIS. OPTICAL PORTION OF THE FLASK WAS MADE OF QUARTZ TO ALLOW TRANSMISSION OF EXCITATION WAVELENGTHS AS SHORT AS 200 NM. EACH GLASS SAMPLE WAS ATTACHED TO THE COLD-FINGER WITH MOLTEN INDIUM. THE DEWAR FLASK WAS EVACUATED TO 0.1 PA (10 TO THE -4TH POWER TORR) BEFORE THE ADDITION OF LIQUID NITROGEN. A MODIFIED AMINO-BOWMAN MODEL 4-8202 SPECTROPHOTOFLUOROMETER WAS USED TO RECORD ALL GLASS LUMINENSCENCE DATA. OF AN INITIAL SAMPLE OF 143, ONLY 21 GLASS SAMPLES WERE FOUND TO BE INDISTINGUISHABLE BY REFRACTIVE INDEX. THIS INDICATED A PROBABILITY OF 0.068 THAT TWO RANDOMLY SELECTED GLASS SAMPLES WILL HAVE THE SAME REFRACTIVE INDEX. FOR 17 OF THE 21 SAMPLES, PHOSPHORESCENCE DATA PERMITTED DISCRIMINATION. THE REMAINING SAMPLES, WHILE NOT INDIVIDUALLY DISTINGUISHABLE, WERE TWO DISSIMILAR PAIRS. FOR ALL SAMPLES THE PHOSPHORESCENCE SPECTRUM CONSISTS OF TWO PRINCIPAL BANDS WITH VARIABLE INTENSITIES THAT PEAK IN THE RED AND THE GREEN. THE RATIOS OF THE INTENSITIES OF THESE BANDS FOR VARIOUS EXCITATION WAVELENGTHS WERE REQUIRED CRITERIA FOR DISCRIMINATION AMONG SAMPLES. THE MEASUREMENTS CAN BE MADE WITHOUT DESTRUCTION OF THE SAMPLES AND REQUIRE 1 MG SAMPLES IF STANDARD INSTRUMENTS ARE USED. DATA DEMONSTRATE THAT THE DISCRIMINATION OF WINDOW GLASS SAMPLES CAN BE ENHANCED BY PHOSPHORESCENCE ANALYSIS. ADDITIONAL RESEARCH WILL BE NEEDED TO IDENTIFY THE IMPURITIES RESPONSIBLE FOR PHOSPHORESCENCE AND TO DETERMINE THE FACTORS THAT CONTROL THE SPECTRAL PROPERTIES OF THESE IMPURITIES. RESEARCH ON OTHER TYPES OF GLASS IS ALSO RECOMMENDED. REFERENCES ARE INCLUDED. (JAP)