Dark matter, memories and chimeras
[This is my Lab Report column for the forthcoming (February) issue of Prospect.]
The universe has had its first X-ray, and the skeleton looks much as expected. It is made of mysterious dark matter, which accounts for five sixths of the matter in the universe. It has been mapped out in a project called the Cosmic Evolution Survey (Cosmos), which pieced together more than 500 images from the Hubble space telescope to trace out the fine filaments of dark matter threading through deep space. These act like flypaper for ordinary matter, concentrating it into stars and galaxies.
In effect, Cosmos has revealed the hidden framework behind the entire visible world. Among other things, it’s another reason to be thankful that Nasa was persuaded not to leave the ageing Hubble to decay, rather than risk another shuttle-borne service mission.
Dark matter can’t be seen directly—that’s why it is dark—and isn’t made from any known subatomic particles. But we know it’s there because of the gravitational tug it exerts on visible matter. It was first postulated to explain why galaxies don’t fly apart as they rotate—but not everyone was convinced it existed until a study last year revealed it apparently balancing an otherwise lopsided spread of visible matter in a collision between two galaxies.
The Cosmos team mapped the distribution of dark matter over a big chunk of the universe by looking at the way it bends light from very distant galaxies behind it. The densest clumps of dark matter, where filaments intersect, mostly match the positions of visible galaxies and gas, as they should. But there are a few regions of dark matter without visible matter and vice versa—bones without flesh, and flesh without bones, so to speak. That wasn’t predicted and can’t easily be explained. Might there be entire galaxies made only of dark matter? It seems more likely that the discrepancies are just errors in the data, which depends on a tour-de-force of astronomical measurement.
It remains something of a scandal that we know nearly nothing about the stuff these cosmic bones are made from (our ignorance of the “dark energy” that outweighs dark matter by a factor of three is even more profound). But that’s modern cosmology for you: the harder you look, the more puzzling it gets.
Another well-hidden scandal lies in the guts of modern computers. Imagine that your piano had to be retuned from scratch every time you wanted to play, or your television set had to be retuned to the broadcast frequencies whenever it was switched on. This is the way computers work: you turn on and wait for minutes as the working memory, or random access memory (RAM), relearns how to work by reading the “manual” held on the hard disk. Actually, it’s even worse. The RAM then forgets everything you tell it within a microsecond—or would do if its memory cells were not refreshed thousands of times a second. That’s why, if the power turns off unexpectedly, all your unsaved data is lost.
It’s also why the RAM sucks all the juice from laptop batteries within hours.
All this could be changed with a technology that is already commercially available but still too expensive to go into your computer. Non-volatile RAM is a memory that, once loaded with data, holds onto it indefinitely without needing any power. With such a memory, a computer wouldn’t need a hard disk at all, and would be ready to use the moment you switched on. Various forms of non-volatile RAM are being developed, the most advanced being ferroelectric RAM (FeRAM), in which the memory elements are a little like magnets with poles oriented by electric rather than magnetic fields.
FeRAM is big in east Asia—the leaders include Samsung, Matsushita and Fujitsu. There’s even a small FeRAM in the Sony Playstation 2. But big, computer-style memories remain too expensive to make this way, and so FeRAM is currently relegated to low-tech applications such as smart cards for Japanese railways. They can be read from six inches away, however, so you don’t even need to get the card out of your bag.
Dr Moreau visions
The UK Human Fertilisation and Embryo Authority recently concluded that “there needs to be a full and proper public debate and consultation” about whether it should licence the creation of animal-human chimeric cells, for example with human DNA housed by nonhuman cells. The HFEA wisely avoided a snap decision in response to the opposition to such cells in a government white paper. The issue, as it says, is “far from black and white.” But the white paper’s recommendations seem to suffer from visions of Dr Moreau. Xenotransplants of animal organs and tissues into human patients have been happening for years, but the popular idea, peddled by the human genome project, that genes represent the sacred core of humanity has apparently created a sense that they are untouchable.
While there are ethical difficulties with chimeric cells, the medical benefits could be huge. Stem cells can be cultivated from embryos, but human eggs are hard to acquire for this. Animal eggs carrying human DNA would be more readily available, and could usher in all kinds of stem-cell therapies. The stem cells wouldn’t be “half-animal” in any meaningful sense, any more than a molecule taken from a human cell is different from an identical one taken from an animal. The HFEA aims to complete its consultation in the autumn. UK legislation in this area has been sensible and permissive so far, but stem-cell scientists are nervous that this time the yuck factor is about to kick in.