Leonhard's family studies, although richly descriptive, may be methodologically outdated by present-day criteria. However, his conjectures have found support in a twin study,70 in which the highest MZ/DZ concordance rate differential was obtained with 'unsystematic' schizophrenia (88.9% MZ, 25.0% DZ) and the lowest with cycloid psychoses (38.5% MZ, 36.4% DZ), consistent with the prediction about the role of genetic factors in the aetiology of the different psychoses. A genome scan of 12 German multiplex pedigrees with periodic catatonia (one of the 'unsystematic' schizophrenias with a 26.9% recurrence risk71), revealed significant linkage on chromosome 15q15, and suggestive evidence on 22q13.72, 73 The follow-up of the 15q finding has so far excluded the nicotinic acetylcholine receptor alpha7 subunit gene, the zinc transporter SLC30A4, and the NOTCH4 gene.74, 75, 76 The putative 22q13 locus contains the MLC1 (WKL1) gene, coding for a cation channel expressed exclusively in the brain and causing, in its mutated form, a severe neurodegenerative disorder (megaencephalic leukoencephalopathy). A rare missense mutation in this gene, cosegregating with periodic catatonia in a single pedigree was detected, but not confirmed in an independent sample.77, 78 Although the jury is still out, the apparent clinical homogeneity of the syndrome of periodic catatonia may not be underpinned by genetic homogeneity.