232 Mr. J. F. Heyes on the 



C 2 H 4 . H 2 0," for instance, has fought its way out of the 

 ambiguous company of dotted " molecular aggregations " to 

 be viewed as aldehydrol 



CH 3 



Arr/OH 



0hl \0H* 

 Similarly for acetal and other bodies. Even C 2 H 4 2 . H 2 

 has been suggested to be 



CH 3 



I OH 



0£OH 

 OH 

 though it is not easy to recognize the chemical equality of the 

 first hydroxyl grouping with the others. There are facts 

 which preferentially favour 



CH 3 



H 2 0=-OC— OH; 

 and the recent researches by Ramsay and Young show the 

 existence of the more complex gaseous molecules 

 CH 3 -CO— OH 

 C 4 H 8 4 or |j 



CH 3 — CO— OH 

 (vi.) In the interesting so-called molecular compound dis- 

 covered by Friedel, (OH 3 ) 2 . HC1, a remarkably stable body, 

 we have a most striking case for the tetravalency of oxygen, 

 either as 



H \/ CH3 



ca / x ch 3 ' 



or, as seems preferable, 



/0H 3 /CH 3 



H-C1=0( , or H— Cl==-0( • 



X CH 3 X 0H 3 



(vii.) This stable compound brings us back to the three types 

 HC1, H 2 0, H 3 N with which we started ; but before resuming 

 their consideration, allusion must be made to the polymeri- 

 zation of aldehyds, of cyanates, of the metaphosphates, and 

 other bodies as closely connected with the tetravalency of 

 oxygen*. 



* The stimulus — if I may so term it — to polymerization in these cases 

 seems to he due to the development, in ways that current research are 

 seeking to explain, of the higher valency, or, as it maybe preferably termed, 



the validity of oxygen, or the tervalidity, for instance, of chlorine in hydro- 



