Dis '98 Structure Functions Summary, Part Ii

Recent results presented in the structure functions working group are briefly summarized for the following topics: The theoretical treatment of heavy quarks in structure functions, higher-order corrections for the leading-twist evolution (including small-x resummations), the present status of the proton's parton densities, and the impact of higher twists on determinations of the strong coupling constant. The reader is referred to Part I 1 for accounts of the transition to the photopro-duction region, pion and photon structure results, and high-Q 2 phenomena. 1 Heavy quarks in structure functions The charm structure functions, especially F c 2 , have attracted considerable interest over the past years. Unlike in the fixed-target regime, F c 2 makes up a sizeable fraction, up to about a quarter, of the total F 2 in the HERA small-x region. Despite being suppressed, it contributes significantly as well to the scaling violations in the kinematic range covered by the NMC data. At low scales, Q 2 ≈ m 2 c , F c 2 is uniquely calculated from the light parton densities via the γ * g → c¯ c Bethe-Heitler process and its O(α 2 s) corrections 2 , without invoking the concept of a charm parton distribution (we neglect here a possible intrinsic charm component, which seems to be relevant only at high x). For Q 2 ≫ m 2 c large logarithms appear in the coefficient functions C 2,L , which may require a resummation. At x < 10 −2 these logarithms dominate C 2 already for Q 2 > 20 GeV 2 , but C L only above 10 3 GeV 2 3. Previous leading-order results 4 for this resummation have been extended to higher orders in ref. 5. This leads to a high-Q 2 description in terms of four massless flavours, with the charm distribution uniquely specified by the light parton densities. The problem of the transition between both approaches, i.e., the construction of a variable flavour-number scheme (VFNS) has also been addressed 5,6. There seems to be agreement that a unique construction does not exist, and indeed the prescription of refs. 5,6 differ at non-asymptotic values of Q 2. Note, however, that this ambiguity concerns only the coefficient functions, as the parton evolution can be kept strictly massless without any loss of generality 7 .