hypothetical interstellar molecules HC 19 N. This rather large result is in reason- 

 able agreement with the extrapolation of the successful gas-phase semidetailed 

 calculation to fQ (1CT 11 to 10" 15 ) only if we assume a synthetic bias toward 

 linear, unsaturated species. The number of possible 20-carbon hydrocarbons is 

 quite large and encompasses many different types of skeletal branching. If many 

 of these species have abundances similar to the estimated abundance of HC^ 9 N, 

 then our large cosmic-abundance-determined limit may indeed be appropriate. 

 However, the gas-phase mechanism favors highly unsaturated species. If this 

 mechanism is operative, according to current opinion, then the large abundance 

 of HC 19 N is in line with our extrapolation of the semidetailed calculation. 

 Despite this agreement, we are forced to conclude that there is indeed great 

 uncertainty in trying to estimate the abundances of complex hydrocarbons and, 

 by extension, all other complex organic species based on the limited, present- 

 day observations. It seems reasonable to suggest that our knowledge of the abun- 

 dances of these species in dense interstellar clouds based on microwave observa- 

 tions will increase gradually and incrementally as species slightly larger than 

 those previously seen are detected. 



What are the prospects for observing species more complex than HCjjN? 

 Consider the spectral region currently utilized by radio astronomers; this extends 

 roughly from 1 to 300 GHz, with the lower frequencies labeled microwaves 

 and the higher frequencies labeled millimeter waves. Studies of the radiation 

 physics of emission lines in this spectral region show that for each molecule 

 there is an optimum emission frequency, or frequency of maximum emission 

 intensity, which is a function of both temperature and the size of the molecule. 

 As temperature increases, the optimum emission frequency increases as the 

 square root of temperature. As the size of the molecule increases, the optimum 

 emission frequency decreases dramatically. As an example, for a linear hydrocar- 

 bon with 20 carbon atoms in a cool interstellar cloud (T = 10 K) the optimum 

 emission frequency is about 5 GHz, whereas in a warm cloud (T = 50 K) this 

 frequency rises to about 11 GHz. If the calculations are performed to maximize 

 the observed signal-to-noise ratio, this latter frequency falls to 8 GHz, because 

 the noise level of the receiver increases with frequency. This example shows that 

 the detection of complex molecules via rotational transition frequencies is best 

 attempted at microwave and millimeter rather than submillimeter wavelengths. 



However, there is a fundamental problem in observing complex molecules via 

 rotational transitions at any frequency. As molecules become more complex, 

 their density of rotational states increases, typically (for linear molecules) as the 

 cube of the number of heavy atoms. This means that the number of molecules in 

 any given rotational level decreases dramatically as the number of atoms 

 increases. The intensity of an emission spectral line is proportional to the 

 number of molecules in a given energy level. Hence, the intensity of any spe- 

 cific emission line decreases strongly as the number of atoms increases, because 

 the emission is spread out into many more emission lines. In particular, a linear 



61 



