Hydrogen randomization precedes the formation of M+--H. and M+. -CH; species from the stilbene molecular ion at 15 eV. The carbon atom involved in the M+. -CH; elimination originates randomly from the whole molecule. The [M -151 ion (m/e 165) in the spectra of stilbene and 9,lO-dihydrophenanthrene is produced from a common ion.

+ E spectra derived from the non-decomposing anions of some sulphur compounds are independent of the pressure of the collision gas. The compounds chosen for study contain peaks produced from either ortho effects or skeletal rearrangement fragments in their conuenfionul positive ion spectra. The dissociative + E spectra are either devoid of skeletal rearrangement ions or alternatively contain such peaks in small abundance. In contrast, peaks derived from ortho reactions are present in high abundance.

Peaks due to M -H20 are observed in the mass spectra of aromatic carbonyl compounds which contain an ortho (or peri) methoxy-substituent. This process is of diagnostic value for simple systems, but cannot be extended t o those systems which have a large number of oxygen-containing substituents. The fragmentation has been studied by 2H and l 8 0 labelling. The properties of the M -18 fragment in the spectrum of o-methoxybenzaldehyde correspond to those of a benzofuran radical ion. PREVIOUS studies 293 have shown that M -H,O' ions are of diagnostic value in the mass spectra of aromatic carbonyl compounds containing adjacent (ortho or peri) ethoxy-substituents. Deuterium and l80 labelling studies3 are consistent with the formation of a stable oxonium species by an ' ortho-effect ' of the general type (4 + (b).A logical extension to this work is to determine whether the presence of a methoxy-substituent adjacent to a carbonyl function can be detected by the occurrence of the process M -H,O. This process will not be as pronounced as the M -H,O' process of the ethoxyanalogue, because of the decreased stability of the odd electron species corresponding to b. This work was

The fragmentation patterns in the spectra of 2-aryl-l,3-dithianes and -1,3-dithiolanes have been elucidated by deuterium labelling studies. Ortho effects are observed in the spectra of 2(o-alkoxyphenyl)-l,3-dithianes. The molecular ions of the 1,3-dithianes eliminate SZH. and metastable peaks substantiate these eliminations. The hydrogen involved in this elimination randomises ((2-4, 5 and 6) prior to elimination in the spectrum of 2-phenyl-l,3-dithiane, but originates mainly from C-5 in that of the bis propane-l,3-dithioacetal of terephthaldehyde.

The hydrogen sulphide + carbon reaction has been studied by the flow method in the temperature range 1000-1600" K. Coconut charcoal reacts appreciably at 1OOO" K and the rate increases rapidly as the temperature is raised. For hydrogen sulphide pressures from 0.2 to 0-5 atm at 1200" K the experimental conversions to carbon disulphide (35-7-304 %) and to sulphur (4.0-2.0 %) are within 1 % of the theoretical equilibrium values. Above this temperature the maximum carbon disulphide yields are less than expected, e.g. only 55.5 % compared with a possible 69.7 % at 1600" K. It is probable that the rate of the forward reaction increases more slowly in this range and that some carbon disulphide is lost by the reverse reaction during the cooling of the gas stream. Under these conditions hydrogen sulphide is decomposed significantly into sulphur and hydrogen, but it is not yet clear whether the carbon disulphide is produced directly frcm hydrogen sulphide and carbon or through the intermediate formation of sulphur. Catalysis by sodium carbonate has been observed.Other types of carbon decrease in reactivity towards hydrogen sulphide at 1300" K in the order : beechwood charcoal > coconut charcoal > B.C.U.R.A. reactive chars > coalite > anthracite > coke. Even coke, however, gives a 45.4 % yield of carbon disulphide at 1500" K. Measurements of B.E.T. adsorption, heat of wetting and water uptake indicate that the surface areas and pore radii of these materials cover the ranges 6-856 m* g-1 and 9-167 A. Diffusion either inside or outside the pores is unlikely to be an important rate-determining factor at 1300" K in most cases, so it is understandable that the order of reactivities is broadly that of the surface areas.