During frozen storage at −5 C, dimethylamine (DMA) was produced in the muscle of five gadoid species. The amount was lowest in haddock and increasingly higher in cod, pollock, cusk, and hake. When the dark lateral muscle was removed from the fillets before freezing, formation of DMA during frozen storage was either inhibited or greatly reduced. Under the same storage conditions no DMA was produced in the muscle of halibut, plaice, redfish, or wolffish.

The use of the Nash test, in conjunction with TCA extraction, for measuring formaldehyde in fish muscle was made more quantitative. This was done by means of a "recovery factor" which took into consideration the percent of formaldehyde added to the muscle extracted by the TCA solution. The average recovery from 15 different samples of cod muscle was 51.3% with an SD of 5.6. Because of differences between species in capacity to bind formaldehyde, it would appear that a different "recovery factor" may be required for each species of fish. As heating muscle increases its ability to bind formaldehyde, recovery factors developed for use with raw muscle are not applicable to the same muscle after it has been cooked. The percentage of added formaldehyde that was recovered varied with variations in the procedures used in preparing the muscle, making the extract, and carrying out the Nash test.

In cod fillets undergoing deterioration during frozen storage, the dimethylamine content increases (and not the trimethylamine content as previously reported by us). There was no evidence to show an accumulation of dimethylamine in the muscle of frozen scallops, lobster, or shrimp that were purposely held at relatively high storage temperatures. It is suggested that for fish of the family Gadidae dimethylamine might be used as a measure of "frozen-storage deterioration" in much the same way as trimethylamine has been used as a measure of microbial spoilage in the unfrozen fish.

Addition of formaldehyde to fresh cod muscle, to give concentrations of 10 to 200 ppm, brought about marked decreases in the extractable protein content during holding periods of 24 hr or less at 0 C. Similar levels of formaldehyde, produced during frozen storage of gadoid (Atlantic cod, Gadus morhua, pollock, Pollachius virens, cusk, Brosme brosme, and silver hake, Merluccius bilinearis), fillets at −5 C, brought about similar reductions in the extractable proteins. Comparative tests showed that, in the concentrations normally encountered in deteriorating frozen gadoid fillets, formaldehyde was a much more active protein-insolubilizing agent than free fatty acid. It is evident that in these protein changes more than one mechanism is involved. Observed species-differences in the extent to which fish proteins became insolubilized during storage appeared to be related to presence or absence of these different mechanisms. The more rapid and more extensive denaturation of most gadoid fillets in frozen storage than of fillets of nongadoid species appears to be directly related to the presence of muscle enzyme of the former group that is capable of producing formaldehyde from trimethylamine oxide, which is absent in the muscle of the nongadoid species so far tested.

During frozen storage at −18 and −25 C the lipids in cod muscle did not undergo oxidation, as indicated by thiobarbituric acid values and odours. In fact they underwent a marked decrease in the ease with which they were oxidized by added Cu++, Fe++, or hemoglobin. This change preceded the protein denaturation that occurs in stored frozen muscle and appeared to be directly related to the formation of free fatty acids in the muscle. A similar change in the sensitivity to metal-induced oxidations could be produced in fresh, unfrozen muscle by the addition of mixed fatty acids prepared from several marine lipids.The addition of four pure saturated fatty acids had little or no effect on the development of rancidity in muscle, either in the presence or absence of added metal catalysts. Fish muscle appears to exert a protective action against the oxidation of added linolenic or linoleic acids. Unlike the mixed marine fatty acids, pure linoleic and linolenic acids did not suppress the development of metal-induced rancidities in fish muscle lipids.