48 SYNTHESIS BY SUNLIGHT IN RELATIONSHIP 



to this a saturated solution of ammonium carbonate is added 

 gradually with shaking so long as the precipitate first formed 

 continues to dissolve. The solution is then dialysed for some days, 

 ten days or more, until the reaction for chloride becomes very faint. 

 The solution so obtained is of a deep reddish-brown colour, even in 

 a 1 per cent, solution, and if converted back into the crystalloidal 

 form by boiling with a drop or two of acid the change is remarkable, 

 to a pale lemon-yellow colour. At the dilutions used in the experi- 

 ments below the colour scarcely shows when reduced to the crystal- 

 loidal condition, but in the colloidal condition, even at this dilution, 

 the solutions possess a deep sherry colour. The dilute solution in an 

 ordinary small test-tube absorbs the blue of the spectrum completely, 

 as shown by a pocket spectroscope. When examined by light trans- 

 mitted from a " Uviol " mercury lamp, the solution viewed directly 

 transmits a yellowish- green light, and at the sides, reflected from the 

 glass surfaces, there is a deep green fluorescence, which reminds one 

 strongly of the fluorescence of a strong solution of chlorophyll. 



The colloidal iron solution so prepared is readily coagulated by 

 boiling, and is most sensitive to added crystalloids ; it is thrown out 

 by 1 part in 1000 of ammonium carbonate, and a mere trace of deci- 

 normal caustic soda throws it completely down. It is in a deli- 

 cately reactive, meta-stable condition, which reminds one forcibly, 

 as it did Graham fifty years ago, of the proteins and the constituents 

 of living cells. 



When set up in the " Uviol " apparatus and the transmitted 

 light observed with a spectroscope, it is seen that the bright lines of 

 the mercury arc spectrum in the blue and violet have entirely dis- 

 appeared, and the only ones now visible are those of the red, orange, 

 and green. An examination of the solar spectrum shows complete 

 absorption of all higher wave-lengths than green. 



There is this difference between the solar and the mercury arc 

 light absorption, that in the former there is a continuous spectrum 

 absorbed from green onward, while in the mercury arc spectrum 

 the absorption is that of three sets of wave-lengths, one at the 

 junction of green and blue, the other far over in the blue, and the 

 third in the violet portion of the visible spectrum. We have not 

 hitherto been able to observe the absorption of the ultra- visible 

 rays. The light energy from these definite wave-lengths of the 

 spectrum seems, however, from the results recorded below, to be 

 very effective for the particular synthesis under consideration. 



