X-ray diffraction studies indicate that addition of the elements of alkoxymercuric salts occurs with inversion at one of the vicinal carbon atoms in alkenes (1). By contrast our studies with 1 , 4-~~-methylene-As-cyclohexene-2, 3-14dicarboxylic acid (I) and with 1 , 4-~~-ethylene-A3-cyclohexene-2, 3-JJ-dicarboxylic ester (XV1II)l seem to show that addition in these compounds occurs directly, Le. without inversion at either carbon atom comprising the alkene linkage.The configuration of 1, 4-TT-methylene-A5-cyclohexene-2, 3-JL-dicarboxylic acid (I) has been demonstrated by Alder and Stein (2) who prepared it by hydrolysis of the adduct from cyclopentadiene and maleic anhydride. When I is boiled with aqueous mercuric acetate the mercuric salt which is formed immediately is slowly transformed into an oxymercurial which is either 5-L-hydroxyanhydro -[6-J-hydroxymercuri-2-J-carboxy] -1 ,4 -TTmethylenecyclohexsne-3-J-carboxylic acid (11) or 5-J-hydroxy-6-~-hydroxymercuri-2-J-carboxy-1,4-~~methylene-cyclohexane-3-J-carboxylic acid lactone (IIa). When this compound is treated with acetic anhydride an equivalent of water is lost and 5-4-hydroxyanhydro -[6-J-hydroxymercuri-Z-J-carboxy] -1,4 -TTmethylenecyclohexane -3-L-carboxylic acid, y-lactone (111) is formed. Alternatively I11 may be formed directly from I by treatment with mercuric acetate in either 2,5-dioxahexane or methanol. It may be noted that this lactonized carboxymercurial is similar in the type with the esterified acetoxymercurial obtained by the addition of mercuric acetate to cyclohexene. Presumably both I1 and I11 are preceded in the system by the mercuric salt IV since the initial precipitate in the reaction system forms mercuric oxide when treated with alkali whereas the final precipitate only dissolves in this reagent.The existence of the lactonized anhydro-hydroxymercurial I11 would seem to require that oxymercuration occurs directly (without inversion) but this conclusion depends on the existence of the mercurial as the monomer 111. No evidence has previously been presented to show that anhydro-hydroxymercuri-The use of the prefixes "exo" and "endo" for specification of configuration in these bridged cycloalkanes and cycloalkenes becomes confusing when more than two substituents must be located, and this situation cannot be ameliorated by introduction of the alreadyabused terms "cis" and "trans". Therefore we have elected in this discussion to use arrows which, if the largest cycloalkane ring is considered as flattened, will designate "up" or "down" in respect of this ring. According to our present nomenclature endo-norbornene-2,3-dicarboxylic acid (I) becomes 1,4-~~-methylene-~6-cyclohexene-2,3-J~-dicarboxylic acid, while disodium 5-endo-hydroxy-6-exo-hydroxymercurinorbornane-2,3-endodicarboxylate (XVI) becomes 5-~-hydroxy-6-~-hydroxymercuri-l,4-f~-methylenecyclohexane-2,3-iJ-dicarboxylate. Likewise endo-5-chloromercuri-6-hydroxy[2.2.2]bicyclooctane-~rans-2,3-dicarboxylic acid y-lactone (XXV) becomes 5-J-hydroxy-6-~-chloromercuri-2-~-carboxy-1,4-ff-eth...

No abstract