A new analytical method for the determination of thesulfur species (SO
2, H
2S, S
80) in volcanic gases is proposed by revising, updating, and improving previousmethods. The most significant advantages of the proposedprocedure can briefly be summarized, as follows: (i) thereaction among sulfur species stops during the gassampling by using preevacuated thorion-tapped vials withpurified 0.15M Cd(OH)
2 in 4 M NaOH to favor theprecipitation of H
2S as CdS; (ii) all the sulfur species (SO
2,H
2S, S
80) are analyzed by ion chromatography, afterconversion to SO
4, which allows the detection limit to belowered significantly with respect to the previous studies;(iii) appropriate aliquots from intermediate steps may beused to determine other species commonly present involcanic gases such as CO
2, HCl, HF, HBr, HI, and soforth; (iv) determination of all the other gas species is notjeopardized by the proposed method, i.e., one single vialcan be used for analyzing the full chemical compositionof a volcanic gas with the exception of NH
3. Statisticalparameters calculated from gas sampling data at the F5crater fumarole in
Vulcano Island (Aeolian Islands, southern Italy), suggest that the standard error of mean (
s/
n)is higher for S (0.10), followed by SO
2, H
2S, and CO
2(0.04, 0.038, and 0.028, respectively). SO
2 shows thehigher variation coefficient (12.1%) followed by H
2S, S,and CO
2 (5.7, 1.5, and 0.8%, respectively). Furthermore,if the time dependence of sampling is taken into account,the measured values, instead of fluctuating in a randommanner, tend to follow systematic patterns, out of statistical control, possibly suggesting a sort of natural fluctuation of the volcanic system. Other crater fumaroles fromvolcanic systems located in different geodynamical areas(Hawaii, USA, El Chichon, Mexico, Poas, Costa Rica) havebeen analyzed as well.