This book is an eye opener and a mind expander. As a science book for the general reader, I give it four stars; this is because Lynn Margulis is a maverick within biology today, and not all that she says is generally accepted science, and because its basic organizational principle, the division of living organisms into five kingdoms, is somewhat out of date. (Since the book came out in 1995, genetic data has made kingdoms subservient to the three "domains" of archaea, bacteria, and eukaryotes.) For those with a more extensive background in biology, give it five stars for its capacity to open up broad new perspectives and to offer illuminating new details.
Lynn Margulis does not serve up any final answer to her title's question. There are a couple of ongoing themes: that wherever there is life, there is what she calls "autopoiesis", the definition of a boundary between self and other, together with the absorption and expenditure of free energy to maintain the self. (A process, as she notes, which not only doesn't violate the second law of thermodynamics, but actually accelerates the rate at which overall entropy increases.) A second theme is, that life's self-organization goes on at progressively higher levels of integration: from cells to colonies and to symbiotic unions that make one complex cell out of several; from complex cells to multicellular animals, plants, and fungi; from multicellular beings to societies and ecosystems; from ecosystems to the biosphere. Margulis believes that biology impoverishes itself by insisting, as Steven Jay Gould does, that evolution has no "direction," simply because no master designer is imposing a direction on it from outside.
But ultimately, according to Margulis, life can't be defined because it keeps on defining itself, and coming up with new definitions. It is better to step back from our skewed view of primates and vertebrates as the most typical living beings, and look at the broad range of specific ways of being alive that evolution and symbiosis have produced so far. So her book is largely organized into chapters describing the history and nature of each of the five kingdoms, in chronological order: bacteria, protoctists (single celled organisms with nuclei), animals, fungi, and plants.
Did you know that it takes two eggs and three sperm to make a flowering plant? That the cell walls of fungi are made from the same material as lobster skin? That photosynthesis evolved independently several times among bacteria?
As stimulating as the book's parade of facts about the wild profusion of ways of being alive is its parade of speculations: that cilia and flagella are the remnants of spirochete bacteria which took up residence in other cells; that the first fertilization event was the result of a failed attempt at cannibalism; that species-jumping genes from fungi taught flowering plants how to make fruit. Another startling hypothesis surfaces every few pages. Some of these speculations are seriously defended, some are tossed out for what they're worth, but they're all fascinating.
I do wish the footnotes had been more extensive. It would be good, for example, to read up on Kwang Jeon's ameba experiments, in which those amebas that didn't die of a bacterial infection wound up incorporating the bacterial genes - and after a few years of lab reproduction, became unable to survive without them. That experiment is the most vivid support Margulis gives for her thesis that genetic material continues to be regularly exchanged across kingdom boundaries; but it's not among the items for which she chose to provide citations.