### Irrational Numbers in Leftins

Finding the square root of a leftin suffers from many of the same problems as
division with leftins. You're not guaranteed a certain number of results or even
any results at all. For a more thorough look at square roots, see the
Square Roots section.

We will approach extracting roots much the way we approached division: simply
try digits from right to left to see what works.

#### Irrational Numbers in *L*_{0}

We will start by trying to find the square root of 17 base 2.
We are looking for b*L*_{0}
such that b^{2} = 17 (or 1001 base 2). The right-most digit of b must
be 1 if it's square is going to end with a 1 (by AP1).

When we try the second digit, we find b can be ...01 or ...11 since the square
of these numbers is ...01. The third digit gives us the four possibilities
...001, ...101, ...011 and ...111. But we find that the fourth digit doesn't
allow two of these answers, though we still end up with four possible values.

At each stage from here on out, we find that the next digit eliminates two of the
previous four possibilities but ends up giving us four possibilities itself.
This means that the left-most digit is ambiguous until we add another digit.
Thus we have two square roots of 17.

...0110011110100110010011011101001

...1001100001011001101100100010111

Both square roots of 17 are irrational numbers, so we know we can represent at least
two irrational numbers in *L*_{f} . All rational numbers can be
represented in *L*_{f} ; is it possible all irrationals are
present? The answer is no.

b*L*^{2} ,
such that b^{2} = 10 base 2.

**Proof:**

By SQ1,
if b exists, it must be an element of *L*_{0} .
If we try squaring all the two-digit base 2 numbers, we find the following:

...00^{2} = ...00

...01^{2} = ...01

...10^{2} = ...00

...11^{2} = ...01

By AP1 we can see that 2
doesn't exist in *L*^{2}.

This brings us to the conclusion that some irrationals are equivalent to
members of *L*_{f} , but not all.

*I* *L*_{f}
.

*I* *L*_{f} .

Back to Index
© Scott Sherman 1999