NUCLEIC ACIDS, DECODING, ETC. 167 



that are indistinguishable by the analysis of their total purine and 

 pyrimidine composition. Nucleic acids having identical total 

 nucleotide contents, though derived from different organisms, 

 differ widely in sequence characteristics. (2) A large proportion 

 of the pyrimidines — and therefore also of the purines — occurs as 

 tracts of three or more in a row. All-thymine sequences of five or 

 even longer occur in many instances, but cytosine has so far not 

 been seen in units larger than triplets. (3) The nucleotide ar- 

 rangement appears not to be random. This could be taken as an 

 indication that a DNA chain could indeed represent a meaningful 

 text. But is a text without a reader still a text? "What is the sound 

 of one hand clapping?" asked the Zen buddhists. 



Which are, one might ask, some of the immediate problems in 

 the chemistry of the nucleic acids? First, a better understanding 

 of what is meant by a DNA molecule. Is the apparent heteroge- 

 neity of DNA ostensible or real? Is there one giant molecule of 

 DNA per chromosome — of a molecular weight of hundreds of 

 milHons — or are there many, different both chemically and func- 

 tionally? 



Secondly, if we really deal with giant chains, do they consist only 

 of polynucleotides or are many tracts, each consisting of several 

 hundreds of nucleotides, held together by bridges that are not of 

 nucleotide character? The fact that DNA is depolymerized by 

 alkali to smaller polynucleotides, but still of respectable size, 

 would seem to indicate the second possibility. I have been inter- 

 ested in this problem for years, but with inconspicuous success. 

 There is an insuperable difficulty in demonstrating, among 

 hundreds or thousands of components of one kind, one of another. 

 It is very hard to distinguish between a trace constituent and a 

 trace contamination. One could think of many types of possible 

 bridging compounds. A derivative of a hydroxyamino acid, such as 

 phosphoserine, or a bifunctional amino acid, e.g., glutamic acid, 

 could form a diester, or an oUgopeptide could do the same. 

 If something of the order of glycerophosphate, ribophosphate, or 

 even a ribonucleotide, were inserted into the chain at infrequent, 

 but possibly predetermined, intervals, regions of reactivity dif- 



