wave-lengths of light (400 to 700 millimicrons) 

 using the Beckman DU spectrophotometer with 

 reflectance attachment. To obtain a diffuse or 

 matt surface, and eliminate factors such as 

 gloss and directionalcolor, the meat was forced 

 through a 16-mesh stainless steel screen and 

 packed heaping full into a 1 1 /4 -inch aluminum 

 planchet. A glass plate was pressed firmly on 

 to the meat until a uniform and smooth surface 

 was produced. The filled planchet was intro- 

 duced into the reflectance attachment of the 

 spectrophotometer and measurement was made 

 in comparison with a pure high-fired alumina 

 disk covered by a similar glass plate. 



A typical series of such normal spectral 

 reflection curves is given in figure 1. The re- 

 producibility of the individual readings was 

 within an average of five-tenths percent reflect- 

 ance unit for tuna flesh, although for color 

 stable materials, such as mixtures of powdered 

 colored glasses, no variation was discernible. 

 Reflectance curves nneasured on replicate sam- 

 ples from the same region of the fish loin gave 

 an average precision of 1.5 percent reflectance 

 units. Reflection from cooked tuna meat is 

 generally high, being highest in the red end. 

 The existence of the Soret absorption at about 

 4 10 millimicrons points to heme protein pig- 

 ments as the source of color. It should be 

 noted that absorption "peaks" lead to decreased 

 reflectance, and so appear as troughs on the 

 percent reflectance curves used in this report. 



100 



UJ 80 

 o 



y 60 



Li. 

 LlI 



LlI 

 O 



cr 



LlJ 

 Q. 



40 



20 



1 r 



-EARLIER RUN 

 LATER RUN 



No attempt was made to fix the specification of 

 this color on Munsell or similar systems since 

 comparison, rather than standardization, was 

 involved in the study. The complicating factor 

 of oxidative browning was present, and caused 

 a general lowering in reflectance with time 

 after cooking (fig. la). At times a heightening 

 of color (increased reflectance) was noted on 

 frozen storage in nnetal foil wrapping (fig. lb), 

 but this may have been due to unusual sample 

 variation. Further work to eliminate or control 

 this additional browning variable is in process. 



Samples similarly judged by experts to be 

 green were prepared and measured. A typical 

 reflectance curve is shown in figure 2 with a 

 non-browned normal curve shown for compari- 

 son. It will be observed that curves for green 

 and normal are sonnewhat similar save for a 

 general decrease in brightness and the indica- 

 tion of a pignnent absorbing in the blue (440-500 

 nnillimicrons) wave-lengths for the green meat. 

 It seems anomalous that the perceived effect is 

 an apparent green since less blue-green reflec- 

 tion is involved. The effect may be due to the 

 higher proportionate red-to-blue reflectance 

 for normal than for green tuna, since the ratio 

 of reflectance of red (640 millimicrons) to blue 

 (410 millimicrons) was 4. 7 for normal and 3, 5 

 for green flesh. The effect can be emphasized 

 by reference to figure 3, which shows the re- 

 flectance curve of tuna meat that haa been 

 soaked in a green dye solution (malachite green) 



T 



1 — r 



T 



1 — r 



/ 



/ 



/ ,^ 



/ 



/ 



OM^^ 



J L 



I 



J L 



••(b)- 



-(b)- 

 -(a)- 

 •<a)- 



I 



/ 



± 



J \ I 



400 440 480 520 560 600 



WAVE LENGTH (MjJ) 



640 



680 



Figure 1. --Reflectance of normal cooked tuna meat: (a) an unwrapped sample in v^ich reflectance 

 decreased with time, (b) a metal foil wrapped sample in which reflectance increased with time. 



