detected in the infrared, can be seen only in selected hot regions such as star- 

 forming areas. Vibrational spectra can sometimes be seen in absorption if there 

 exists a suitable continuum infrared source such as hot dust surrounding a star 

 embedded deep in the cloud. Nevertheless, practical difficulties in the use of 

 ground-based infrared astronomy, such as atmospheric interference, have hereto- 

 fore relegated this technique to a distant second place behind radio astronomy in 

 the observation of gaseous interstellar molecules. 



The two important in situ mechanisms for the production of complex gas- 

 phase molecules in dense interstellar clouds are synthesis from atoms and smaller 

 molecules by reactions on the surface of dust grains followed by desorption into 

 the gas or by reactions in the gas phase. Both of these mechanisms have been 

 studied in some detail. In general, it has been difficult to determine whether gas- 

 phase reactions or grain-surface reactions are more important, although the 

 former are currently favored. Quantitative, time-dependent treatments that con- 

 tain calculated abundances of large molecules have been undertaken for both the 

 grain-surface mechanism and for the gas-phase mechanism. These calculations 

 involve sizable numbers of molecules and reactions, but terminate at species of 

 rather small size in the context of biology. It is possible to simplify the compu- 

 tational problems involved in these models and still obtain agreement with 

 observation. In particular, a "semidetailed" steady-state treatment of the gas- 

 phase chemistry in the dense interstellar cloud TMC-1 can account for the abun- 

 dances of organic hydrocarbons as complex as methyl acetylene (C 3 H 4 ) or the 

 radical C 4 H. Unfortunately, extension of even this treatment to larger molecules 

 is hindered by a lack of rate-coefficient data for important reactions. As these 

 data are obtained in the laboratory, the semidetailed treatment can be extended 

 to include larger molecules. A still more approximate solution of the kinetic 

 equations has been utilized to estimate how efficient gas-phase reactions are in 

 producing complex interstellar hydrocarbons. In this approach, hydrocarbons 

 through 12 carbon atoms in size have been considered. 



A severe upper limit to the abundances of complex hydrocarbons in the gas 

 phase can be obtained from cosmic-abundance arguments. Assume that the total 

 fractional hydrocarbon abundance (with respect to the gas density) in the gas 

 phase of dense interstellar clouds is about 10~ 4 . This is approximately the total 

 fractional carbon abundance and is itself probably an upper limit. Also assume 

 that the total abundance of all hydrocarbons with n carbon atoms is a factor of 

 a times the abundance of all hydrocarbons with /7-1 carbon atoms where 

 < a < 1. This assumption is clearly an idealization, but it is hard to imagine 

 the abundance of all hydrocarbons with n carbon atoms increasing as n increases 

 for a significant range of n. According to a variety of theoretical treatments, the 

 fractional abundance of 1 -carbon hydrocarbons, fQ x (mainly methane), is 10" 6 . 

 Then a is equal to 0.99, and complex hydrocarbons possibly can have high 

 abundances. For example, the fractional abundance of 12-carbon hydrocarbons 

 would have an upper limit of 9.0X1 0~ 7 . Thus, the use of cosmic-abundance 



59 



