Researchers at Rensaelaer Polytechnic Institute have challenged the long standing belief that the deterioration of concrete exposed to freeze-thaw cycles was due to the expansion of the water present in all concrete. For years concrete engineers have reasoned that as water froze and expanded in volume, it set up disruptive forces within the concrete mass. These enlarged spaces in turn, thought the engineers, permitted greater amounts of water to enter the concrete which, upon freezing, further enlarged the void structure of the concrete mass. But information developed by the Rensselaer researcher seems to indicate that in reality the smaller the amount of ice that is formed within the concrete mass, the greater may be the likelihood of deterioration. Their findings apply principally to aggregates rather than hardened cement paste matrix. Thirty-two samples of carbonate rocks from the state of New York were used in the tests. They were divided into two types: 19 that had proven to be frost-sensitive in the field- that is, those that deteriorated rapidly in natural exposures such as quarry faces and 13 that were comparatively frost-resistant. Test showed that relatively all the water in the 19 frost-resistant rocks froze. However, in 8 of the 19 frost-sensitive rocks little or no water froze despite repetition of the freezing up to four times and temperatures as low as minus 40 degrees F. In the remaining 11 frost sensitive rocks, usually less than one-half of the water contained in the rock froze. In addition, it was discovered that freezing did not take place progressively, but rather took place in a single freezing pulse, between 19.4 degrees and 10.4 degrees F. The Rensselaer researchers concluded than that the water in rocks that resistant freezing is absorbed. Further tests done with formamide concluded the deterioration of the rocks was the result of reorienting of polar molecules on adsorption surfaces in a warming and cooling environment.