Current studies in this Laboratory involve the thermal decomposition of certain allylic trimethylammonium hydroxides. The observation that quite appreciable amounts of dimethylamine and saturated carbonylic compounds are formed in these reactions has led to the belief that these quaternary bases rearrange to some extent during the pyrolysis, giving vinylic trimethylammonium bases. Although it is conceivable that the carbonylic products might be formed by the rearrangement of allylic alcohols derived from the unrearranged quaternary bases, the only readily understood mechanism by which dimethylamine could be formed involves the hydrolysis of a vinyldimethylamine, formed by an Sn2 attack by hydroxyl ion upon a methyl group of the vinyltrimethylammonium ion. The products of such a reaction would be methanol and a vinyldimethylamine, which would be hydrolyzed (in the dilute acid commonly used to absorb the bases formed in Hofmann degradations) to dimethylamine and a saturated carbonylic compound.1 2 A thorough examination of the products formed in the thermal decomposition of the simplest allylic trimethylammonium hydroxide (I) was considered worth while to substantiate these findings and postulates.The pyrolysis of I has not been studied for over half a century,3 and then only cursorily. In an obscure publication which appeared in 1886 and has since been overlooked or discredited, Bono (3) heated allyltrimethylammonium iodide with potassium hydroxide and obtained an aldehyde, CeHioO, whose boiling point, 130-135°( as he pointed out) corresponded to that of a-methyI-/3-ethylacrolein (IV), the only aldehyde of that formula known at that time; since our findings (see below) substantiate Bono's identification of this product,4 the question mark appearing in Beilstein's account (5) of his work should be deleted. A short time