news and views

the SO2 flux measurements by COSPEC, ‘order of release’ of volcanic gases from mag- allow for remote determination of HCl, HF mas as related to their relative solubility has

and SiF4, and of changes in the flux of several been a long-term guiding idea of work in this gas species with excellent time resolution. area14,15 — there is now hope that intrusion of (Incidentally, the authors1,2 use different fresh, gas-rich magma into magma bodies, reporting units, which can be related by tons suggested by some interpretations of volcanic 100 YEARS AGO dǁ1 ǂ 0.012 ǃ kg sǁ1). gas data16, might be unambiguously detected While we were taking sympathetic Application of these new methods will by remote sensing and used to increase the breaths with the insatiable shag, the add much to our knowledge of accuracy of forecasts. 8 latter reappeared — yet again with a 15- volcano–atmosphere interactions by giving William I. Rose and Gregg J. S. Bluth are in the inch eel. Four 15-inch eels — all us improved data on many species, particu- Department of Geological Engineering and Sciences, swallowed alive — within the space of larly the halogen compounds. For example, Michigan Technological University, Houghton,

about four minutes! … Would he bring up the study of the ratios HCl:SO2 and HCl:HF Michigan 49931, USA. another? Yes, there he was again with can provide information about the rise and e-mails: [email protected] another 15-inch eel! A very vigorous eel degassing of magma, and is thus a means of [email protected] 2 — just like the others in size and eruption monitoring and prediction . Like- 1. Love, S. P., Goff, F., Counce, D., Siebe, C. & Delgado, H. Nature appearance, and swallowed in the same wise, quantitative measurements of these 396, 563–567 (1998). 2. Francis, P., Burton, M. R. & Oppenheimer, C. Nature 396, manner, after about 30 seconds’ species can be used to constrain geochemical 567–570 (1998). resistance. This made five eels. The models of magma equilibrium and oxida- 3. Symonds, R. B., Rose, W. I., Bluth, G. J. S. & Gerlach, T. Rev. question now arose as to what would be tion state, and of the interactions among the Mineral. 30, 1–66 (1994). the end of this bird. Was he going to die solid, liquid and gas phases3. 4. Baxter, P. J. & Gresham, A. J. Volcanol. Geotherm. Res. 77, 1,2 325–338 (1997). the death of King Henry I before our These advances will also challenge our 5. Stoiber, R. E. & Jepson, A. Science 182, 577–578 (1973). eyes? ... To make a long story short, we understanding of volcanic gas measure- 6. Krueger, A. J. Science 220, 1377–1379 (1983). 7. Bluth, G. J. S. et al. Nature 366, 327–329 (1993). counted twelve eels! — all stout 15- ments. In the case of SO2, geochemists have 8. Rose, W. I. et al. Bull. Volcanol. 48, 181–188 (1986). inchers. The twelfth seemed, perhaps, had great difficulty in interpreting gas flux 9. Goff, F. et al. Geol. Soc. Am. Bull. 110, 695–710 (1998). rather feebler than the others, but it data in a consistent way, and many volcanoes 10.Gerlach, T. M. et al. J. Geophys. Res. 102, 8005–8019 (1997). nearly got away. H.R.H. now seemed to have been reported to release ‘excess’ sul- 11.Andres, R. J. et al. J. Volcanol. Geotherm. Res. 46, 323–329 reflect that this last misadventure was a phur11,12; that is, they release more sulphur (1991). 12.Gerlach, T. M. et al. in Fire and Mud (eds Newhall, C. G. & warning, swallowed his twelfth, and took than expected from study of the mother rock. Punongbayan, R. S.) 415–434 (Univ. Washington Press, Seattle, flight... . There is, of course, only one This uncertainty can apparently now be clari- 1995). explanation of all this; the twelve eels fied by a better appreciation of how the oxy- 13.Scaillet, B. et al. J. Geophys. Res. 103, 23937–23949 (1998). were one and the same eel … . The gen fugacity (effective molecular pressure) in 14.Greenland, L. P. et al. Geochim. Cosmochim. Acta 49, 125–129 (1985). peculiar procedure of ejecting the prey melts influences the separation of SO2 (ref. 15.Gerlach, T. M. & Graeber, E. J. Nature 313, 273–277 (1986). under water appears very remarkable. 13). An understanding of the pattern or 16.Harris, D. M. & Rose, W. I. Bull. Volcanol. 58, 163–174 (1996). From Nature 8 December 1898. Molecular 50 YEARS AGO Since last August there have been reports A hydrogen-producing in the Press of a crisis among biologists in the U.S.S.R. The crisis culminated in a decree from the Præsidium of the T. Martin Embley and William Martin Academy of Sciences ... . This decree dismisses a number of prominent biologists from their posts, closes two ark, damp (and sometimes smelly) studied in anaerobic for 25 years2. famous laboratories, and removes places harbour some of nature’s most have often been suspected orthodox geneticists from committees… . Dcurious eukaryotes. The likes of of stemming from the same endosymbiotic The Lenin Academy of Agriculture swamps and intestines are swarming with bacterium that gave rise to mitochondria. Sciences, of which body Lysenko has mostly single-celled eukaryotes () But now, Akhmanova et al. report a been president for ten years, held a that, like all cells, must produce ATP to sur- that has its own genome, Conference during July 31–August 7 of vive. Yet these places lack enough oxygen to directly betraying its endosymbiotic past. this year… Lysenko’s address followed sustain ATP synthesis as it occurs in textbook Cells of Nyctotherus (which do not grow familiar lines ... . It denies the mitochondria like our own. Some protists in culture and have to be carefully microma- chromosome theory of heredity. It possess no mitochondria, surviving from nipulated from cockroach hindguts) contain arraigns several Soviet biologists because anaerobic fermentation in the cytosol. hydrogenosomes that can be labelled by anti- their work is inconsistent with Marxist Others have quite odd mitochondria that bodies against DNA. The authors found that ideology and is sterile of practical results harbour anaerobic ATP-producing path- the cell produces a ribosomal RNA which, … . The objective of Lysenko’s attack on ways. On page 527 of this issue, Akhmanova although not proven by in situ hybridization Mendelism is not so much its false et al.1 report a gem of such an odd mitochon- to localize to the , bears all the ideology, but rather its impotence. By drion in the ciliate Nyctotherus ovalis. sequence characteristics expected of ciliate insisting on the stability of the germ The ciliate lives in the suffocatingly oxy- mitochondria. Plus, it may look like a mito- plasm Mendelism “condemns practical gen-poor confines of cockroach intestines, chondrion, but this DNA-bearing organelle workers to fruitless waiting”. The where it helps the insect to digest cellulose. is unquestionably a hydrogenosome because inheritance of acquired characters is a Instead of consuming oxygen, Nyctotherus’s it produces hydrogen — Akhmanova and doctrine necessary for the progress of mitochondrion has the bizarre property of colleagues found hydrogen-consuming3 Soviet agriculture. excreting hydrogen as a by-product of ATP methanogenic endosymbionts inside the From Nature 11 December 1948. synthesis. Similar hydrogen-generating cells of Nyctotherus. Finally, Nyctotherus — hydrogenosomes — have been expresses a nuclear-encoded gene for a NATURE | VOL 396 | 10 DECEMBER 1998 | www.nature.com Nature © Macmillan Publishers Ltd 1998 517 news and views

Electron Excreted Example Which ATP-producing acceptor end product organelles occur.....in which eukaryotes? Today - H - - Trichomonads O2 H2O Many

7 N - - - Diplomonads Fumarate Succinate Flatworm NO - N O Fungus8 3 2

e - H - M Percolozoa

NO - NO - Ciliate9 8

m 3

i 2

t

l

a + 1

c H H Ciliate

i 2

g - ? M M Euglenozoa o

l Figure 2 Some electron acceptors, and the o

e corresponding reduced end-products associated G with energy metabolism in mitochondria. N - - - Entamoeba the oxidative breakdown of reduced carbon compounds. Electrons removed during the - H M M Ciliates oxidation process must be dumped onto an electron-accepting compound (an accep- Nyctotherus tor) that can be excreted from the cell. Oth- Origin of eukaryotes erwise, ATP production — and life — comes to a halt. Our mitochondria use oxygen as the acceptor and excrete water. But the mito- - - ? M Stramenopiles chondria of anaerobic eukaryotes must resort to compounds other than oxygen (Fig. 2). Some use organic acceptors such 7 8,9 - - M M Metazoa as fumarate , some use nitrate . The mito- chondria of Nyctotherus1, like other hydro- genosomes2,4, simply transfer the electrons onto protons, producing hydrogen. N H M M Fungi Why are anaerobic ATP-producing path- ways so widespread in eukaryotes2–4,7–10? There are two popular hypotheses for their - - - M Plants origin which, in principle, can be tested through gene-by-gene phylogenetic analy- N H M M sis. One view is that the genes were acquired Spectrum of Typical mitochondria by eukaryotes through horizontal gene ecology and types of Anaerobic mitochondria transfer from one or more prokaryotic energy metabolism donors, other than the antecedent of mito- found in each group Hydrogenosomes chondria. If this were the case, the genes No ATP-producing organelles for these pathways in anaerobic eukaryotes should trace to different prokaryotic (eubac- Anaerobic Aerobic terial or archaebacterial) sources. An alter- native view is that the eukaryotic genes Figure 1 Summary of ATP-producing organelles across a very limited, and arbitrarily chosen, involved in anaerobic ATP synthesis were spectrum of eukaryotic taxa. The ciliate Nyctotherus ovalis (red star) contains hydrogenosomes, inherited from a single common ancestor of which Akhmanova et al.1 now show contain their own genome. A peculiar group of eukaryotes called mitochondria and hydrogenosomes. These microsporidia are related to fungi6, whence the ‘N’ for that group. Antarctic euglenids contain genes are thought to have been transferred to organelles that may be hydrogenosomes10. Some groups designated ‘N’ may have organelles, but their the host’s chromosomes, because they are role, if any, in energy metabolism is unclear. (Data taken from refs 4, 5 and references therein, and not found in any known mitochondrial also refs 2, 3, 6–10.) genome. In this case, the common ancestor of contemporary eukaryotes would have hydrogenase (an enzyme that makes hydro- widespread among contemporary eukary- acquired, from a facultatively anaerobic gen) that is probably imported into the otes4,6, including groups that are distantly protomitochondrion, a genome’s worth of hydrogenosome with a transit peptide1, as is related in conventional genealogies and, genes for all-purpose survival. These genes the case for most proteins in mitochondria. sometimes, arise from within otherwise were then left to the workings of selection The evolutionary significance of these aerobic groups. Although biologists do not and common descent. From this we can pre- findings is twofold. First, hydrogenosome- agree which groups of eukaryotes might be dict that, across the anaerobic eukaryotes in associated DNA was hitherto a genuine miss- the most primitive, most will find their Fig. 1, each gene should ultimately trace to a ing link4. Such discoveries are rare, and the favourite candidate somewhere in Fig. 1, and single eubacterial source. genes in this DNA are likely to hold exciting most biologists do believe that all eukaryotes We know that the antecedent of mito- surprises. The other, deeper, significance share a single common ancestor. chondria brought with it the ability to emerges when we remind ourselves that cili- The key to understanding hydrogeno- respire oxygen — much of the respiratory ate hydrogenosomes are just the tip of the somes, and why they produce hydrogen, is chain is still encoded in mitochondrial iceberg (Fig. 1). Hydrogenosomes and the energy metabolism4. All known eukaryotes DNA11. But we don’t know what else was in anaerobic (or micro-aerophilic5) lifestyle are generate energy (ATP) by one principle — that symbiont’s biochemical repertoire. It 518 Nature © Macmillan Publishers Ltd 1998 NATURE | VOL 396 | 10 DECEMBER 1998 | www.nature.com news and views

probably existed two billion years ago12 but, e-mail: [email protected] ally much larger than the linear one. Their because it’s no longer around, we cannot William Martin is at the Institut für Genetik der ratio is the Reynolds number Re and, for sequence its genome to find out. However, TU, Spielmannstrasse 7, D-38023 Braunschweig, large Re, equation (1) is impossible to solve. molecular fossils of its lifestyle might be pre- Germany. Moreover, no one in their right mind wants served in nuclear chromosomes, including e-mail: [email protected] the full solution of the turbulent velocity our own, allowing us to piece together the 1. Akhmanova, A. et al. Nature 396, 527–528 (1998). field at all points in space-time. It is the statis- bacterial part of our heritage. 2. Müller, M. J. Gen. Microbiol. 139, 2879–2889 (1993). tical properties of the flow, such as probabili- Is anaerobic energy metabolism in 3. Fenchel, T. & Finlay, B. J. Ecology and Evolution in Anoxic ty distribution functions of velocity or the Worlds (Oxford Univ. Press, 1995). eukaryotes a telling relict of our history, or an 4. Müller, M. in Evolutionary Relationships among Protozoa (eds rate of energy consumption, that are impo8r- oddity with little significance? Eukaryotes Coombs, G. H., Vickermann, K., Sleigh, M. A. & Warren, A.) tant. So, what can we do to understand now regarded as primitive tend to be anaer- 109–132 (Chapman & Hall, London, 1998). turbulence? obes, so these are important questions. The 5. Lloyd, D. Nature 381, 121 (1996). In 1922, looking at the evolution of tur- 6. Embley, T. M. & Hirt, R. P. Curr. Opin. Genet. Dev. 8, 624–629 road to answers leads straight to the genomes (1998). bulent atmospheric conditions, Lewis Fry of eukaryotes from the anoxic world. For sci- 7. Tielens, A. G. M. & Van Hellemond, J. J. Biochim. Biophys. Acta Richardson suggested the ‘cascade picture of entists studying oxygen-shunning eukary- 1365, 71–78 (1998). turbulence’. In this, the largest eddies in a otes with unconventional mitochondria, 8. Kobayashi, M. et al. J. Biol. Chem. 271, 16263–16267 system are created by instabilities of the (1996). with hydrogen-producing mitochondria 9. Finlay, B. J., Span, A. S. W. & Harman, J. M. P. Nature 303, mean streamline flow, as in hurricanes. or with no mitochondria at all, these are 333–336 (1983). These decay giving rise to eddies of roughly exciting times. 10.Simpson, A. G. B., van den Hoff, J., Bernard, C., Burton, H. R. half their size which decay in turn, creating T. Martin Embley is in the Department of Zoology, & Patterson, D. J. Arch. Protistenkd. 147, 213–225 (1997). even smaller third-generation eddies — and 11.Gray, M. et al. Nucleic Acids Res. 26, 865–878 (1998). Natural History Museum, Cromwell Road, London 12.Doolittle, W. F. Proc. Natl Acad. Sci. USA 94, 12751–12753 so on, until the smallest stable eddies lose SW7 5BD, UK. (1997). their energy because of viscous friction that turns it into heat. In high-Re turbulence, Nonlinear physics eddies exist at various scales, from the largest ones at the scale of the system size down to Universality of turbulence the smallest ones at the viscous scale. This situation is called ‘developed turbulence’. Victor S. L’vov In 1941 Andrei Kolmogorov estimated the energy E(R) of eddies of scale R in a unit lsewhere in this issue (Nature 396, and Stokes (1945). The Navier–Stokes equa- volume of developed turbulence to be 552–554; 1998) Bramwell et al. tion for the velocity u(r,t) of fluid at point r ț(ȏ–R)2/3. The assumption here is that the Edescribe the discovery of a new type and time t is simply Newton’s second law for only relevant parameter (besides the obvious of universality in turbulence. It connects this the fluid particle: length scale R and density ț) is a rate of ener- strongly non-equilibrium phenomenon of gy consumption ȏ–. Kolmogorov’s crucial _Ꭿ_u_ 2 classical fluid mechanics with critical Ꭿt + Ꮙu·ᒔᏎu ǃǁᒔp + ȗᒔ u (1) idea was the assumption of universality of phenomena in the thermodynamic equilibri- small-scale motions (on scales R Ӷ L) in um of solids (magnetically ordered crystals). This equates a particle acceleration (the left- developed turbulence. Here, universality Intuitively, hydrodynamic turbulence is hand side) with the forcing due to the gradi- means an independence of the statistical understood as the chaotic motion of fluids ent of the pressure p(r,t) and to the viscous properties of small eddies from the nature of — be it of interstellar dust in spiral galaxies, friction (the term proportional to the kine- the fluid (be it interstellar dust or water), of gaseous planetary atmospheres or of water matic viscosity of a fluid ȗ). independence from the mechanism stirring flowing from a tap (Fig. 1). The length-scales In principle, one has to solve this the flow and independence from the particu- vary from galactic distances of 1016–1018 km, equation to fully understand all turbulent lar geometrical form of the container. through planetary distances of 1,000–10,000 phenomena, but it is a mathematical night- But what about the statistics of large-scale km down to the human scales of 1 mm–10 m mare. If one ignores the nasty nonlinear motions? Most of us used to believe these (in the atmosphere and rivers, as well as in term, (u·ᒔ)u, the mean velocity of typical were non-universal and would depend on the kitchen sink). rivers turns out to be about 106 km hrǁ1, and factors such as the geometry of the system. As Euler’s basic mathematical description a maximum car velocity is found to be 2,000 Bramwell et al. now show, however, there is a of fluid dynamics (1741) was corrected to km hrǁ1, both of which are clearly nonsense. class of turbulent flows for which at least account for viscous friction by Navier (1827) The reason is that the nonlinear term is usu- some characteristics of large-scale turbulent E statistics are universal. Y The authors demonstrate the rate of AR

power consumption P(t) in turbulent flow in O LIBR T CE TELESCOP

an enclosed air gap between two counter- HO A rotating disks. For very large Re the probabil-

ity distribution function QP is Re-indepen- dent when properly rescaled. Their Fig. 1a on VIS/SCIENCE P A ASA, HUBBLE SP page 553 shows plots of ȜPQP versus N -D T (P-Pෆ)/ȜP where Pෆ is the mean value of P and ȜP is the standard deviation of P (a character- AM HAR istic width of the distribution QP). The AD normalized distributions ȜPQP for different Re collapse onto the same curve even though they strongly deviate from Gaussian form for P < Pෆ. Bramwell et al. argue that just two Figure 1 From interstellar space to the kitchen sink. The extreme scales of turbulence range from the factors are important: that the integral Re motion of interstellar dust in the M100 spiral galaxy to water flowing from a tap. is fixed at a constant value and that flow is NATURE | VOL 396 | 10 DECEMBER 1998 | www.nature.com Nature © Macmillan Publishers Ltd 1998 519