arguments and a constant multiplicative factor a do not lead to stringent limits 

 on the abundances of complex hydrocarbons unless the total cosmic abundance 

 resides in the smallest hydrocarbons (i.e., fQ x approaches 10~ 4 ). 



The abundance of hydrocarbons produced by gas-phase reactions has been 

 estimated using the approximate kinetic model mentioned previously. With this 

 model, the total fractional abundance of all 12-carbon-atom hydrocarbons is 

 approximately 1CT 10 , far below the upper limit we have developed here. The 

 semidetailed treatment extends only to 4-carbon-atom hydrocarbons. A crude 

 extrapolation of these data leads to predictions that /£ is somewhere in the 

 range 10~ 9 to 10 _1 x , in good agreement with the former hydrocarbon estimate. 

 So it would appear that despite great uncertainties, approximate and semide- 

 tailed gas-phase theoretical treatments "predict" abundances far below the 

 cosmic-abundance-derived upper limits. The detailed grain-surface treatment is 

 a time-dependent model, as is the detailed gas-phase model. The use of these 

 models to extrapolate complex hydrocarbon abundances will lead to answers 

 that depend on the age of the interstellar cloud. In the gas-phase model, abun- 

 dances of hydrocarbons are maximized at intermediate cloud lifetimes (10 5 to 

 10 6 years) well before steady state is achieved. Extrapolation of calculated 

 abundances at such lifetimes leads to higher estimates than extrapolations based 

 on the semidetailed model. However, an estimate of the minimum time neces- 

 sary to synthesize such complex hydrocarbons via gas-phase reactions yields an 

 answer of approximately 10 7 years, far longer than the time at which the smaller 

 hydrocarbons exist at maximum abundance, thus throwing doubt on the extrap- 

 olations. The grain-surface treatment, on the other hand, shows smaller hydro- 

 carbons with high abundances at a time as large as 10 7 years. Extrapolation of 

 these results yields a value of fc l2 below 10~ 12 . 



It is difficult to conclude much at all from these extrapolations other than 

 that they yield complex hydrocarbon abundances significantly smaller than 

 upper limits based on cosmic-abundance arguments. Increased measurements of 

 important rate coefficients will enable the heretofore successful gas-phase semi- 

 detailed treatment discussed above to be extended to much larger species. 



What do current observations tell us about the abundances of very complex 

 molecules in dense interstellar clouds? A broad infrared feature at 3.4 jum has 

 been interpreted as indicative of complex organic matter on the surfaces of 

 grains. Likewise, other broad infrared features have been interpreted as indica- 

 tive of a wide variety of unresolved polycyclic aromatic hydrocarbons in the gas 

 phase. However, our best extrapolations based on distinctive, sharp, gas-phase 

 microwave spectra involve the linear cyanopolyynes, HC /7 N, where n is an odd 

 integer. In the well-studied source TMC-1, HC 3 N, HC S N, HC 7 N, HC 9 N, and 

 HCjjN have been observed. Their abundance drops by a factor of approxi- 

 mately four each time two more carbons are added from n = 5 through n = 9. 

 With a fractional abundance of 10~ s for HC S N, simple extrapolation using 

 this factor yields a fractional abundance of approximately 6X1CT 13 for the 



60 



