roform layer gently (35°C) in an open vial 

 overnight in a hood on a sand bath. The dried 

 extract was reconstituted with 1:9 chloro- 

 form: methanol and chromatographed on a 30 cm 

 Sephadex LH20 column at a flow rate of 3-4 mL/ 

 hour. A fraction was collected by hand every 20 

 minutes, its volume measured, and its fluores- 

 cence determined on the Turner 111 at the same 

 wavelengths used for routine analysis of lipofus- 

 cin. To determine where a retinol peak would ap- 

 pear, we chromatographed a tissue extract to 

 which a commercial preparation (Sigma) of 

 retinol had been added. Also, extracts of larval 

 fish in the pre-eyed stage were compared with 

 extracts of larvae that had pigmented eyes, as- 



suming that the eyed larvae would have more 

 retinol. 



A UV irradiation step had been proposed to 

 degrade retinol where it may interfere with mea- 

 surement of extracted lipofuscin, but Csallany 

 and Ayaz (1976) reported that this procedure was 

 ineffective. We therefore determined the time 

 course of degradation of commercial retinol in 1:9 

 chloroform: methanol by UV irradiation in quartz 

 tubes. 



At the wavelengths we used, retinol was not an 

 important interfering substance in a variety of 

 larval fish and in the adult grunion muscle and 

 adult halibut liver tissues (Fig. 2). There also was 

 little difference between extracts of eyed and pre- 



FlGURE 2. — Chromatograms (fluorescence vs. fraction number) 

 for various fish, compared with fish tissue plus retinol. 



'50- HALIBUT LIVER 



100 



LARVAL GRUNION (whole) 



LARVAL SEABASS (whole) 



20 5 



FRACTION (ml) 



409 



