Evolving Evolution
By Edward Ziff, Israel Rosenfield
From DNA to Diversity: Molecular Genetics and the Evolution of 
Animal Design
by Sean B. Carroll, Jennifer K. Grenier, and Scott D. Weatherbee
Endless Forms Most Beautiful: The New Science of Evo Devo and the 
Making of the 
Animal Kingdom
by Sean B. Carroll

Norton, 350 pp., $25.95
The Plausibility of Life:Resolving Darwin's Dilemma
by Marc W. Kirschner and John C. Gerhart

Yale University Press, 314 pp., $30.00
1.

Despite much recent controversy about the theory of evolution, 
major changes in 
our understanding of evolution over the past twenty years have gone 
virtually 
unnoticed.[1] At the heart of Darwin's theory of evolution is an 
explanation of 
how plants and animals evolved from earlier forms of life that have 
long since 
disappeared; but his theory says nothing about the factors that 
determine the 
shape, color, and size of a particular fish, whale, or butterfly. 
Darwin and his 
contemporaries realized that understanding the evolution of animal 
forms and 
understanding how a fertilized egg develops into a whale, cow, or 
human being 
must be deeply connected; but they didn't know how to make the 
connection.

Surprising discoveries in the 1980s have begun to tell us how an 
embryo develops 
into a mature animal, and these discoveries have radically altered 
our views of 
evolution and of the relation of human beings to all other animals. 
The new 
field of study in which these breakthroughs have been made is 
called Evo Devo, 
short for evolution and development, "development" referring to 
both how an 
embryo grows and how the newborn infant matures into an adult.

Sean Carroll, author of one of the books under review and a 
coauthor of another, 
has made important contributions to the understanding of evolution 
and 
development. From DNA to Diversity, written with two other 
scientists, is the 
second edition of a book that has become a classic for students of 
evolution. 
The title of Carroll's other book, Endless Forms Most Beautiful, 
comes from the 
famous final sentence of The Origin of Species: "There is a simple 
grandeur in 
this view of life... that from so simple an origin, through the 
process of 
gradual selection of infinitesimal changes, endless forms most 
beautiful and 
most wonderful have been evolved."

In 1830, nearly thirty years before Darwin published his book, two 
French 
naturalists?Georges Cuvier and Étienne Geoffroy St. Hilaire?debated 
the 
significance of the anatomical similarities between distantly 
related animals, 
such as the flippers of whales and the wings of bats. Cuvier held 
that form was 
dictated by function: the bat's wing, needed for flying, had a 
separate origin 
from the whale's flipper, needed for swimming.

Geoffroy St. Hilaire opposed this view, arguing that the underlying 
skeletal 
similarities pointed to the existence of a common archetype for 
both flippers 
and wings. While neither man claimed that animal forms could change 
over 
generations, St. Hilaire's archetypal form foreshadowed some recent 
discoveries 
about development and evolution. No doubt this debate was in the 
mind of Charles 
Darwin as he formulated his theories in an attempt to account for 
the origins of 
animal forms.
 

The contemporary Darwinian theory of evolution is based on three 
ideas: natural 
selection, heredity, and variation. Small random 
changes?variations?occur in 
organisms through mutations of genes, and when these changes give 
an organism a 
greater chance of survival, they persist from one generation to the 
next through 
natural selection. That is, organisms with traits that make them 
better adapted 
to the environment they inhabit will have better reproductive 
success than other 
members of the same species that do not possess the advantageous 
traits. In each 
successive generation, then, an ever-larger proportion of the 
species in 
question will possess the mutation that produces the advantageous 
traits. 
"Natural selection," Darwin wrote, "acts solely by accumulating 
successive, 
favorable variations." Evolution in the Darwinian view was gradual: 
"it can act 
only by short and slow steps." And since, in this view, all changes 
are random, 
there are no predetermined directions in which organisms evolve. 
All living 
organisms, Darwin claimed, are descended from one or a few common 
ancestors.

Neither Darwin nor any of his contemporaries knew about the 
workings of 
heredity?how we inherit the eye color of our father or the hair 
color of our 
mother. The work of the Czech monk Gregor Mendel, first published 
in 1865, had 
gone unnoticed in Darwin's day and was only rediscovered around 
1900. Mendel had 
shown that specific traits, such as the color of a pea, or the 
smoothness or 
roughness of its skin, could be inherited independently of one 
another. The new 
science of "genetics," the idea that units called "genes" within 
each cell 
transmit specific traits, such as hair color, from one generation 
to the next, 
began in the first decade of the twentieth century. Studies of 
inherited traits 
in fruit flies in the following decades established convincing 
evidence for 
genes, but they remained invisible. Scientists still didn't know 
how the gene 
made it possible for information to pass from one generation to the 
next, and 
how mutations in genes could, over many generations, lead to a new 
species that 
had a form different from its distant progenitor.

By the 1940s, though the structure of the gene was still unknown, 
scientists had 
introduced the idea of the gene into Darwinian theory. They now 
explained 
evolution as the consequence of small random changes in genes. This 
recasting of 
Darwinian theory was called the Modern Synthesis, following the 
1942 publication 
of Julian Huxley's book Evolution: The Modern Synthesis. The 
neo-Darwinian 
theory incorporated the Mendelian idea of genetics, explaining the 
mechanism of 
inheritance that was unknown to Darwin. The theory, however, did 
not account for 
how particular organisms develop from embryos in the womb to adult 
forms; that 
process, known as embryology, was not discussed.

The neo-Darwinian view was reinforced in 1953, when the double 
helix was 
discovered, showing how genes composed of the nucleic acid DNA 
transmitted 
hereditary characteristics. A molecule of DNA is made up of two 
long strands of 
chemical building blocks called nucleotides, each containing one of 
four bases: 
adenine, thymine, guanine, or cytosine, which are abbreviated A, T, 
G, and C. 
The order of the bases in each strand of DNA determines the 
information in the 
DNA molecule, information we can think of as providing an overall 
plan for 
producing enzymes and other proteins.

A gene was now understood to be a specific sequence of DNA bases. 
Genes vary 
considerably in size, most of them containing between 10,000 and 
20,000 
nucleotides, though they can be much longer or shorter. Each of our 
cells 
carries all our genes, although most genes remain inactive at any 
given moment. 
When a particular gene is activated it is first copied into RNA, a 
nucleic acid 
that carries instructions from DNA for assembling proteins. The 
RNA's 
instructions are then decoded in a process called "translation," 
and proteins, 
including the enzymes essential for cells to function, are 
produced. Proteins in 
turn form some 50 percent of all living cells.

How are particular genes activated? There are, according to recent 
research, as 
many as a hundred trillion cells in the human body, and each cell 
contains 
thousands of different types of molecules that vary considerably in 
size; many 
molecules move about freely inside the cell. All the cells in a 
given individual 
have the same DNA?it is contained in the largest molecules in each 
cell?and 
hence an identical set of genes. Which specific genes are activated 
in a 
particular cell depends, in part, on the cell's location in the 
embryo or the 
adult body. Moreover, the activation of one combination of genes 
will give rise 
to a liver cell, while the activation of another will produce a 
brain cell.

The structure of the double helix made it apparent how changes, or 
mutations, in 
the base sequence of a gene could lead to variations in the 
characteristics of 
an organism; such mutations could, if advantageous, accumulate over 
time. The 
process appeared to confirm Darwin's view that evolution is 
gradual. As he wrote 
in The Origin of Species, nature "can act only by short and slow 
steps. Hence, 
the saying Natura non facit saltum," nature doesn't make sudden 
jumps. The 
standard view, then, was that variation and selection could account 
for how the 
simple organisms of early life evolved into the complex forms of 
the 
contemporary biological world. After Mendel's discoveries had been 
absorbed at 
the beginning of the twentieth century, it was assumed that as 
changes 
accumulated between one species and another, there would be less 
and less 
similarity in the kind and number of their genes. More advanced 
species would 
have many more genes than lower forms of life; and worms, for 
example, would 
have few, if any, genes similar to those of fish, mice, or human 
beings.

Yet it seemed astonishing that random mutations, even over enormous 
periods of 
time, could give rise to the remarkable complexity of an organ such 
as the human 
eye. "To suppose," Darwin wrote in The Origin of Species,

    that the eye with all its inimitable contrivances for adjusting 
the focus to 
different distances, for admitting different amounts of light, and 
for the 
correction of spherical and chromatic aberration, could have been 
formed by 
natural selection, seems, I freely confess, absurd in the highest 
degree....

Nonetheless, he continued:

    Reason tells me, that if numerous gradations from a simple and 
imperfect eye 
to one complex and perfect can be shown to exist, each grade being 
useful to its 
possessor, as is certainly the case; if further, the eye ever 
varies and the 
variations be inherited, as is likewise certainly the case; and if 
such 
variations should be useful to any animal under changing conditions 
of life, 
then the difficulty of believing that a perfect and complex eye 
could be formed 
by natural selection, though insuperable by our imagination, should 
not be 
considered as subversive of the theory.

The neo-Darwinian belief in small mutational changes in DNA 
molecules over 
hundreds of millions of years made the preservation of individual 
genes over 
long periods of time highly unlikely. It was thought that the 
diversity of 
living forms was the consequence of each animal having evolved its 
own unique 
set of genes over millions of years as well. Surely human beings, 
for example, 
would not have the same genes as worms.

Those assumptions were dramatically overturned when the rough draft 
of the human 
genome?the entire set of human genes?was announced in 2001. As it 
turned out, 
human beings have far fewer genes than expected? about 25,000 
rather than the 
60,000 or more that had been predicted. This was about the same 
number as mice 
have, and even the tiny worms called nematodes have about 14,000 
genes. The 
number of genes in a given species, therefore, is not a measure of 
its 
complexity. Why had biologists so overestimated the number of genes 
in the human 
genome? Why is it unnecessary for complex animals such as mammals 
to have ten 
times as many genes as worms?

The answers to these questions were already hinted at more than 
four decades 
ago. At the time it was known that the bacteria E. coli, which 
normally live off 
the sugar glucose, are also capable of producing enzymes that 
digest other 
sugars, such as lactose. But biologists noticed that the bacteria 
only produce 
the enzyme when lactose is present in their immediate environment. 
Scientists 
could not explain how the E. coli somehow "knew" when the 
lactose-digesting 
enzyme would be needed.

In 1961, Jacques Monod and François Jacob discovered that E. coli 
bacteria 
actually have a mechanism that controls the production of the 
enzyme for 
digesting lactose. As unicellular organisms, E. coli bacteria have 
only several 
thousand genes, each of which is made up of a specific sequence of 
DNA. A single 
one of these genes, present in all E. coli, carries in its DNA the 
genetic 
instructions needed to assemble the enzyme that can digest lactose; 
the DNA is 
copied into RNA, which is then "translated" to produce the enzyme 
itself. When 
there is no lactose present in the bacteria's immediate 
environment, the gene is 
switched off: its DNA is not copied into RNA and the enzyme is not 
produced. The 
reason for this, the scientists discovered, is that a protein 
called a repressor 
molecule attaches itself to the DNA site where the copying into RNA 
begins, thus 
blocking off the DNA and preventing the gene from producing the RNA 
responsible 
for the synthesis of the enzyme.

On the other hand, when the E. coli bacteria encounter lactose, the 
lactose 
binds itself to this repressor molecule, causing the repressor to 
be detached 
from the DNA site. This unblocks the DNA, allowing the gene to be 
copied into 
RNA, and produce the enzyme that can digest lactose. In other 
words, the 
repressor molecule acts as a switch that controls the gene's 
production of the 
enzyme. Since only a fraction of the total number of genes present 
in an 
organism are expressed, or turned on, at any given time, Monod and 
Jacob 
conjectured that other genes must be similarly turned on and off.

Although they had not yet found systematic evidence to support 
these ideas, the 
discovery of the repressor molecule allowed the two scientists to 
form a 
powerful new hypothesis about how genes function. As Jacob recently 
wrote, in a 
brief description of the new hypothesis:

    It proposed a model to explain one of the oldest problems in 
biology: in 
organisms made up of millions, even billions of cells, every cell 
possesses a 
complete set of genes; how, then, is it that all the genes do not 
function in 
the same way in all tissues? That the nerve cells do not use the 
same genes as 
the muscle cells or the liver cells? In short, [we] presented a new 
view of the 
genetic landscape.

The deeper significance of the Monod-Jacob model of gene function, 
and its 
implications for the nature of evolution, became apparent with the 
new field of 
embryo research that arose almost twenty years later.
2.

In 1894, the English biologist William Bateson challenged Darwin's 
view that 
evolution was gradual. He published Materials for the Study of 
Variation, a 
catalog of abnormalities he had observed in insects and animals in 
which one 
body part was replaced with another. He described, for example, a 
mutant fly 
with a leg instead of an antenna on its head, and mutant frogs and 
humans with 
extra vertebrae. The abnormalities Bateson discovered resisted 
explanation for 
much of the twentieth century. But in the late 1970s, studies by 
Edward Lewis at 
the California Institute of Technology, Christiana 
Nüsslein-Vollhard and Eric 
Wieschaus in Germany, and others began to reveal that the 
abnormalities were 
caused by mutations of a special set of genes in fruit fly embryos 
that 
controlled development of the fly's body and the distribution of 
its attached 
appendages. Very similar genes, exercising similar controls, were 
subsequently 
found in nematodes, flies, fish, mice, and human beings.

What they and others discovered were genes that regulate the 
development of the 
embryo and exert control over other genes by mechanisms analogous 
to that of the 
repressor molecule studied by Monod and Jacob. Eight of these 
controlling genes, 
called Hox genes, are found in virtually all animals ?worms, mice, 
and human 
beings? and they have existed for more than half a billion 
years.[2] Fruit flies 
and worms have only one set of eight Hox genes; fish and mammals 
(including 
mice, elephants, and humans) have four sets. Each set of Hox genes 
in fish and 
mammals is remarkably similar to the eight Hox genes found in fruit 
flies and 
worms. This discovery showed that very similar genes control both 
embryological 
and later development in virtually all insects and animals. (See 
Figure 1.)

To understand what Hox genes do, scientists needed to observe the 
activity of 
the genes in the developing embryos of flies and mice. Using new 
technologies 
that allowed them to observe under a microscope the locations of 
the Hox 
proteins in these embryos, they were able to identify an overall 
pattern of how 
Hox genes behave. A newly fertilized fly egg looks like a tiny 
football: one 
end, or pole, will eventually become the head region; the other end 
will become 
the tail region. These and other divisions of the embryo in later 
development 
actually followed the switching on and off of the Hox genes in 
different parts 
of the embryo. (See Figure 2.)

The mechanism that causes the Hox genes to behave in this way is 
initiated by 
the release of proteins from the cells of the mother's body across 
the newly 
fertilized embryo. These proteins control the activities of the Hox 
genes and 
are released in varying concentrations, causing Hox genes to 
produce Hox 
proteins in specific places. As the embryo grows, the production of 
Hox proteins 
divides the embryo into a series of segments, or distinct regions, 
from which 
subsequent development occurs. Other genes are then activated 
within each 
segment, a finer division of the embryo is established, and wings, 
antennae, and 
other body parts are formed. In general, scientists reasoned, Hox 
genes 
establish the basic division of the embryo into distinct 
compartments, and each 
compartment, in turn, establishes the regions of the embryo where 
development of 
specific body parts and functions takes place. Still, the details 
of the 
mechanisms that a cell uses to establish its position in the embryo 
remain 
incomplete.

In fact, Nüsslein-Vollhard and Wieschaus found that within the 
fly's embryo 
there was an overall pattern in which genes were turned on or off; 
and they saw 
in this pattern the overall body plan for the full-grown fly. In 
other words, 
the activity of the Hox genes, including the formation of 
compartments within 
the embryo and the control of other genes that guide development, 
provided a 
system of organization that dictated the final adult form of the 
fly.

The presence of a body plan in the genome, whether of a fly, a 
whale, or a 
human, was unexpected by embryologists. Previously, most of them 
did not think 
that development of embryos was controlled by genes; they had 
assumed that the 
different parts of developing embryos were determined by physical 
interaction 
between neighboring cells and that there was no overall division of 
the embryo 
according to a genetic plan. Experiments had shown, for example, 
that removing 
developing wing tissue from one part of an embryo and implanting it 
elsewhere 
still gave rise to a wing, although an abnormal one. Scientists 
attributed the 
abnormality to the effects of the neighboring cells in the embryo. 
This was 
wrong. In fact it was caused by the disruption of the body plan.

This new understanding of Hox genes was aided by the discovery that 
the proteins 
produced by these genes function in a way that is analogous to 
Monod and Jacob's 
repressor molecule. Specifically, although they have different 
properties, all 
Hox proteins contain a molecular structure that makes them attach 
to DNA sites 
that control genes. This meant that Hox proteins, like the 
repressor molecule, 
act as switches that turn neighboring genes on and off.[3]

Hox genes, as Carroll explains, are in fact one of many kinds of 
genes that 
direct embryo development by a mechanism of switches. One example 
that is not a 
Hox gene is the gene that controls the development of the eye in 
fruit flies. If 
this gene (called Pax 6) is damaged when it mutates, the fly fails 
to develop 
eyes. We now know from the experiments described in Carroll's book 
that the Pax 
6 gene is also found in butterflies, mice, and humans. Indeed, Pax 
6 genes are 
interchangeable. The Pax 6 gene from a fly will turn on genes that 
make eyes in 
mice, and the Pax 6 gene from a mouse will turn on genes that make 
eyes in 
flies.

It had long been assumed that eyes had evolved independently in 
different 
species. The structures of mammalian eyes and insect eyes are very 
different and 
it would seem most unlikely that they had followed a similar 
evolutionary path. 
Mammalian eyes, for example, have a single lens that focuses an 
image on the 
retina. Insects have eyes with many tube-like structures, each tube 
having its 
own lens and retina. Yet the discovery of the Pax 6 gene gives us 
reason to 
believe that the evolution of the eye in all the animals followed 
related, and 
to some extent common, paths, though we cannot completely exclude 
the 
possibility that each kind of eye evolved following completely 
independent 
pathways. In addition to the Pax 6 gene, genes have been found that 
control the 
genes responsible for the development of the different kinds of 
"hearts"?or 
mechanisms that pump blood?whether in fruit flies or humans, again 
suggesting 
similar evolutionary pathways. Indeed the development of legs, 
wings, arms, 
fins, and other fish and marine animal appendages are all under the 
control of 
virtually identical genes and, as with the Pax genes, in many cases 
are 
interchangeable.

These findings strongly support the Darwinian view that animals 
descend from one 
or a few ancestors. However, contrary to the previously accepted 
neo-Darwinian 
view, the same findings showed that different animal forms are not 
primarily a 
function of distinct gene pools that have evolved over millions of 
years. How 
then do similar collections of genes create the enormous diversity 
of living 
forms? In Sean Carroll's view, what creates diversity is the 
patterns in which 
genes are turned "on" and "off." The different appendages found in 
centipedes, 
fruit flies, lobsters, and brine shrimp are created by varying 
combinations of 
Hox gene activity in the developing insect or crustacean embryo.

"Switches," Carroll emphasizes, "enable the same...genes to be used 
differently 
in different animals" [his italics]. In other words, a Hox gene 
produces a 
protein that binds to the DNA's sites where genes copy into RNA and 
can thus 
turn genes "on" or "off."

This has an important consequence for evolution: mutations in Hox 
genes will 
affect the ways in which they act within the embryo, thereby 
altering the 
proteins' functions as switches. When the proteins' functions are 
changed, in 
turn, this causes them to control genes that are needed for 
development of a 
specific physical trait in new ways. In this view, evolution is 
largely the 
consequence of random mutations in genetic switches. Genes remain 
intact, but 
under new patterns of control. Their function is altered. 
Complexity and variety 
are created, at least in part, by combining the activities of old 
genes in new 
ways. Carroll's view?what we might call the switches 
hypothesis?emphasizes the 
importance of changes in patterns of turning genes on and off 
rather than 
changes in the genes themselves. However, even the most ardent 
supporter of the 
switches hypothesis would admit that not only Hox genes but other 
genes change 
as well. But the contribution of these changes to evolution is far 
less than we 
had previously believed.

In fact, vertebrates?reptiles, birds, chickens, mice, pythons, and 
humans? do 
have more genes than insects, though far fewer than had been 
expected before the 
human genome was revealed. The increase in the number of genes in 
these animals 
is partly responsible for their complexity and diversity. But as 
Carroll notes, 
"frogs and snakes, dinosaurs and ostriches, giraffes and whales, 
have evolved 
around a similar set of four Hox gene clusters. So again, the mere 
number of Hox 
genes does not tell us how these forms evolved." The diversity of 
these animals 
comes from changes in the ways genes are turned on and off.

For example, though the giraffe has a long neck, it has seven 
cervical 
vertebrae, the same number as humans, whales, and all other 
mammals. Hox genes 
control this number, but they may also control cell proliferation 
and 
consequently the size of the vertebrae. The giraffe's larger 
vertebrae may have 
developed because of mutations in the Hox genes controlling the 
size of 
vertebrae. Giraffes with large vertebrae and longer necks could 
feed off tall 
trees and were consequently selected over other giraffes. Changes 
in gene 
regulation, not new genes, gave rise to the long-necked giraffe.

Evolution, then, depends on new patterns of gene regulation rather 
than the 
creation of new genes. Indeed, it is not meaningful to talk about 
the function 
of a single gene in isolation. Genes only function in the context 
of the 
organism. There is no single gene for an eye, a limb, or language, 
much less 
such tendencies as homosexuality. Genes function in relation to 
other genes and 
intercellular signals, much as words vary in meaning and function 
depending on 
the way they are used in sentences and the contexts in which they 
are spoken. It 
is the combinations of gene activity, which may be different in 
different 
species, that create the form of the organism. "We can begin to 
think of 
individual groups?insects, spiders, and centipedes, or birds, 
mammals, and 
reptiles, as well as their long extinct fossil relatives?not so 
much in terms of 
their uniqueness, but as variations on a common theme," Carroll 
writes. And 
surprising, too, is the evidence that all animals, from worms to 
humans, 
probably descend from one or a few primitive bacteria. Darwin would 
have been 
pleased to discover molecular evidence for his "common descent."
3.

A powerful new theory adds considerable weight to this view, 
putting Carroll's 
work in a larger perspective. In The Plausibility of Life, Marc W. 
Kirschner and 
John C. Gerhart take a broader view than Carroll's on the questions 
of 
development and evolutionary biology. They agree that Hox genes 
make an 
important contribution to the mechanisms of evolution, but they 
argue that there 
are a number of other fundamental properties of organisms that give 
direction to 
evolution. The weakness of Darwinian theory?and one that has been 
seized upon by 
secular critics of evolutionary theory?is its failure to explain 
how the gene 
determines the observable traits of the organism. From an 
evolutionary point of 
view, how can complex organs such as eyes, arms, or wings evolve 
over long 
periods of time? What about the intermediary forms?

The Darwinian view was that early evolutionary forms of arms, legs, 
or wings 
might have initially served other purposes (insect legs, for 
example, might have 
evolved from gills their ancestors used for respiration). Such 
transformations 
of purpose are certainly important in evolution, Kirschner and 
Gerhart argue, 
but there must be other mechanisms at work as well. Concerning the 
human eye, 
for example: How is it possible for the different parts of an eye 
to evolve 
simultaneously?the lens, the iris, the retina, along with the blood 
vessels 
necessary for supplying the eye with oxygen and nutrition as well 
as the nerves 
that must receive signals from the retina and send signals to the 
muscles of the 
eye? Could these precise nerve and vascular networks be created by 
gradual 
random changes in genes over long periods of time, as Darwin 
claimed? Similarly, 
how can random mutations and natural selection create not only the 
necessary 
muscles and bone that make up the arm, for example, but organize 
the blood 
supply and nerves so that, after hundreds of thousands of years, an 
animal 
evolves with functioning arms, legs, and eyes? The Darwinian view 
that 
developing organs can serve different purposes at different times 
seems 
incomplete at best.

Darwin thought that at any given time variations in the forms of 
organisms were 
purely random. This is true of the neo-Darwinian view as well. 
However, recent 
research has shown that even though mutations are random, the 
effects of a 
mutation will be restricted, and may alter only one part or trait 
of an 
organism. A good example of the restricted effects of mutation is 
provided, as 
Kirschner and Gerhart point out, by the body plans created by Hox 
genes. Because 
they are contained within the different compartments of the embryo 
established 
by the body plan, individual parts of an animal can evolve 
independently of each 
other. For example, the lizard has limbs, the python has vestigial 
limbs, and 
the advanced snake has no limbs at all. These variations in limb 
structure have 
evolved without major changes in other parts of the body plan.

This independence means that mutations can occur within a single 
region of an 
embryo that may or may not be beneficial; in any case, fewer of the 
mutations 
will be lethal for the developing organism. In other words, while 
evolution is 
constrained by the body plan created by the Hox genes, this 
constraint gives 
nature a much greater freedom to experiment with variant forms 
through random 
mutations. If there were no body plans with separate parts, most 
variations 
would be lethal to the entire organism and evolution would be much, 
much slower. 
Suppose we wanted to design new windows for airplanes that would 
improve the 
visibility for passengers, resist cabin pressure, and better 
insulate passengers 
from the cold. We would test the new window designs without 
changing their 
positions on the body of the plane. If we had to redesign the 
entire plane every 
time we changed the window design, we would be much slower in 
developing new and 
more efficient planes. Similarly, Hox genes can, through mutations, 
shift the 
pattern of organization within a part of the embryo, allowing 
evolution to 
experiment with new forms, such as wings and longer necks, without 
affecting 
other parts of the embryo.

Kirschner and Gerhart thus place the activity of Hox genes inside 
the embryo in 
a broader perspective. They agree that Hox genes are important in 
organizing the 
embryo into discrete parts, a process they call the invisible 
anatomy (it is 
only visible with the aid of recent technology). But they argue 
that the 
function of Hox genes is only one of a number of "core processes" 
that act as 
constraints on evolution. The storage of genetic information in DNA 
and the 
mechanisms for translating that information in the synthesis of 
proteins are 
examples of core processes. Other kinds of core processes that are 
used by cells 
include biochemical mechanisms, such as the digestion of nutrients 
by enzymes. 
These mechanisms were established at an early stage in evolution 
and are still 
used in human cells, worms, and bacteria. Because of these core 
processes, 
natural selection is presented with a variety of forms that are 
more likely to 
succeed than if there were no constraints on variation at all. 
Should a new 
advantageous process arise by mutation, it can be incorporated into 
the 
functional repertory of the organism, and it is then inherited over 
generations.

Another kind of core process that can, by constraining development, 
create forms 
that are more likely to succeed is what Kirschner and Gerhart call 
"exploratory 
behavior," such as the method used by ants when foraging for food. 
Ants leave 
their nest and take random paths. As they move about they secrete a 
chemical 
substance called a pheromone that leaves a scent along the path 
they are 
following. If an ant fails to find food it will eventually return 
to the nest, 
using the pheromones it has deposited to guide it back to the nest. 
However, an 
ant that finds food will deposit more pheromones as it returns to 
the nest. This 
will reinforce the scent of the trail that led to food and other 
ants will now 
follow the reinforced trail.

Nonetheless, not all the ants will follow the successful trail. 
Some ants will 
set out on random paths in search of other sources of food and if 
successful 
they, too, will establish paths for subsequent ants. Eventually, 
the ants will 
have established a detailed map of paths to food sources. An 
observer might 
think that the ants are using a map supplied by an intelligent 
designer of food 
distribution. However, what appears to be a carefully laid out 
mapping of 
pathways to food supplies is really just a consequence of a series 
of random 
searches.

Other exploratory processes are important for the development of 
the vascular 
and nervous systems in a growing embryo. While the details of the 
individual 
processes vary considerably, the guiding principles are similar to 
those of ant 
foraging: just as the ants randomly explore the terrain around 
their nest, 
capillary vessels sprout off from the larger blood vessels and 
randomly explore 
the surrounding tissues for the signals coming from cells deprived 
of oxygen; 
they then can bring blood containing oxygen to the cells. And just 
as contact 
with food makes the ant reinforce the path that led to the food, 
the sprouting 
capillary vessels establish permanent contacts whenever they 
encounter tissue 
with oxygen-deprived cells. Similarly, fine nerve endings extend 
themselves 
randomly, establishing stable connections between nerves and 
muscles whenever 
they receive electrical and chemical signals coming from muscles.

Hence, unlike the eye or hand, whose forms follow from the body 
plan that is 
programmed by Hox genes in the developing embryo, the apparently 
well-designed 
and integrated nervous and vascular systems that serve such organs 
do not 
require predetermined paths and wirings. Darwin's view that small 
simultaneous 
changes would give rise to organs as complex as the eye is in 
principle true, 
but in need of modification. It is the very constraints created by 
the Hox genes 
and the other core processes (e.g., the exploratory behavior of 
capillaries and 
nerve endings) that permit complex designs to emerge over a 
relatively short 
period of time from a biological point of view (hundreds of 
thousands of years, 
or perhaps even less). As Carroll and Kirschner and Gerhart 
observe, some 
alteration of genes is still necessary if the changes are to be 
passed on from 
one generation to another. But the genetic alterations are 
considerably simpler 
and fewer in number than had been formerly imagined.

While Carroll argues?a claim that is at the heart of Evo Devo?that 
embryological 
development gives us the deepest clues to the mechanisms of 
evolution, Kirschner 
and Gerhart move beyond embryology to show that metabolic and 
physiological 
processes are also critical to evolutionary change. Their approach, 
which they 
call the theory of "facilitated variation," attempts to show how 
the regulation 
of genes inside the embryo, as described by Carroll, is part of a 
larger set of 
processes that allow organisms to experiment with evolution in a 
tightly 
controlled way. According to this theory, the mutations, or 
variations, needed 
to drive evolutionary change can occur with little disruption 
either to the 
basic organization of an organism or to the core processes that 
make its cells 
function.

We now have a far deeper understanding of evolution than even a 
decade ago.[4] 
And although our knowledge is still incomplete, our new 
understanding, as the 
books under review admirably show, has opened the way toward a 
comprehensive 
account of evolution and has supplied solid answers to the critics 
of 
evolutionary theory.
Notes

[1] For their critical readings and comments on this article, we 
would like to 
thank David Botstein, Nathaniel Heintz, Luisa Hirschbein, David 
Ish- Horowicz, 
and Richard Lewontin, none of whom should be held accountable for 
this text, for 
which we take full responsibility.

[2] No one group found all of the Hox genes. Lewis started his work 
in the late 
1950s, earlier than Nüsslein-Vollhard and Wieschaus, but much of 
the work of the 
two groups was contemporaneous. Both groups realized that 
development was under 
genetic control. Although Lewis focused on specific Hox genes, 
Nüsslein-Vollhard 
and Wieschaus saw the importance of identifying all of the genes 
that controlled 
the body plan, and they identified most of them.

[3] Hox proteins behave in this way because of a particular genetic 
feature of 
the Hox genes that produce them. Within their genetic composition, 
all eight Hox 
genes possess a nearly identical section of DNA, called a homeobox. 
When Hox 
genes produce Hox proteins, the homeobox region of the Hox genes 
carries the 
genetic information to produce a specific part of these proteins 
called the 
homeodomain. Once the protein is assembled, its homeodomain 
attaches to DNA 
sites that control genes, allowing it to function as a switch.

It is this particular characteristic that gave the Hox genes their 
name. In 
homage to Bateson, the identical sections of DNA were called 
"homeoboxes," since 
they were present in genes that, when mutated, resulted in 
Bateson's monsters, 
or "homeotic" mutants. The term "Hox gene" is a whimsical 
combination of 
"homeotic" and "homeobox."

[4] While classical genetics relies on changes in the order of the 
bases in the 
DNA strands, other heritable changes have been discovered?changes 
in the 
chemical makeup of the bases that control gene activities, for 
example?and are 
increasingly recognized as important in development and evolution. 
For example, 
the base cytosine might undergo a chemical modification involving 
the addition 
of a single carbon atom.