294, REPORTS ON THE STATE OF SCIENCE. 
Kay and Perkin! find that commercial magnesium powder sometimes 
does not interact with isopropyl iodide in ethereal solution. When, how- 
ever, the magnesium is left in contact with an ethereal solution of methyl 
iodide until the action is vigorous, then washed thoroughly with ether, 
and at once mixed with the ethereal solution of isopropyl iodide, the 
action proceeds satisfactorily. 
(7) Importance of preventing oxidation of the reagent.—It should be 
noted that Bouveault and Bodroux have found that the Grignard reagent 
is acted on by oxygen to form primary alcohols or phenols, so that not 
only should the reagent be protected from moisture and carbon dioxide, 
but it should not be left unnecessarily long in contact with air before 
being applied. Sometimes, indeed, it is necessary to prepare the reagent 
in an atmosphere of hydrogen.” 
Methods of applying the Grignard Reagent. 
As a general rule the compound to be acted on is dissolved in 
anhydroas ether and the solution either added to the Grignard reagent 
or vice versd. It is convenient to perform this operation by means of a 
siphon and to take precautions that moisture and carbon dioxide are 
excluded. As a rule, the action requires to be regulated, otherwise it 
may become too violent. When the addition is complete, sometimes the 
mixture is heated. Powdered ice is then gradually added, then mineral 
acid. The product of the action is generally in the ethereal solution. 
The following description, taken at random, is typical. It is the 
method used by Ullmannand Miinzhuber ® for the preparation of tripheny]- 
carbinol from methyl benzoate. 
A mixture of magnesium (7°2 grams), freshly distilled bromobenzene 
(57°2 grams, 25 mol.), anhydrous ether (60-70 grams), and a trace of 
iodine is placed in a large round-bottomed flask provided with a reflux 
condenser and then gently warmed. As soon as the brownish liquid is 
decolorised and is boiling rapidly it is cooled by immersing the flask in 
ice-cold water ; the action is regulated so that the boiling continues to 
be vigorous. After twenty minutes the bulk of the magnesium is dis- 
solved ; it is practically all gone after the boiling has been continued for 
half an hour longer. The flask containing the reagent is then placed in 
ice-cold water, and by means of a dropping-funnel a solution of 20 grams 
of methyl benzoate (freshly distilled) in an equal volume of anhydrous 
ether gradually added. The action is vigorous. After the addition of 
the ester the mixture is heated to boiling for about an hour, ice-cold 
water gradually added, and then dilute sulphuric acid. The ether is 
expelled from the ethereal solution and the product distilled in steam to 
remove unchanged bromobenzene and a small amount of diphenyl ; the 
triphenylcarbinol remains in the flask. Yield of crude product, 33 grams 
(87 per cent.) ; when crystallised from alcohol 28 grams of pure product 
are obtained. 
Of course this method of working cannot always be adopted. Water 
in some cases must not be added to the product of the action (compare 
Kipping’s work on silicon compounds). Again, in the action of carbon 
dioxide and other gases on the Grignard reagent special methods have to 
1 Trans. Chem. Soc., 1905, 87, 1081. 
2 Tschitschibabin, Ber., 1905, 38, 561; Tschelinzeff, ibid., 1906, 89, 773. 
* Tbid., 1903, 38, 406. 
