*I have had problems with Google Books and   one of my main gripes is the searchability of their books.  
When searching a text at the Google site, I usually get SOME accurate results, but most often I don't get 
ALL the pages.  When I then download the PDF book and do a search, my Adobe Reader tells me there are 
NO results which is very frustrating!  As a result of no results, I search for a book at other sites and only 
download a Google Book if there are no other options. 

I found the following article that addresses Google Book issues:

SOURCE: Medieval Genealogy

LINK:
http://www.medievalgenealogy.org.uk/google.shtml
    

Google Book Search - Hints and Tips

Google Book Search is already a very useful resource, but in some respects it is still not very user-friendly. 
I have tried to collect here a few useful hints and tips for finding and using material. 

[Thanks to Al Magary for generously providing some of the information below.] 

-Contents 
-Searching 
-Navigating 
-Problems accessing material outside the U.S.A. 
-Submit a link 
-Contents

The Google Book Search database contains three kinds of books: 

Books considered to be out of copyright, or "public domain". 
Books scanned in libraries but still in copyright. 
Books submitted by publishers. 
(Note that Google applies different copyright criteria according to which country the visitor is in. Many 
books accessible to visitors in the U.S.A. are not accessible to those elsewhere. See Problems accessing 
material outside the U.S.A..) 

The texts of all these books can be searched, but how much is shown depends on the status of the book. Out 
of copyright books can be viewed in their entirety. For books submitted by publishers, a page or two either 
side of each search result can be viewed (this may require the visitor to register before viewing the pages). 
For other books, at most only "snippets" - typically two or three lines containing each search result - can be 
seen. 

Note that although this is (I think) how the system is meant to work, there are numerous anomalies. Many 
books do not appear to be accessible in any form, although they appear in lists of search results, and some 
books can be viewed only as "snippets" even though they are clearly out of copyright. 

Searching
Google allows searching of the texts of books in its database, using either a basic or an advanced search 
form. 

Currently the advanced search form, in addition to the usual Google search options, allows the visitor to 
supply search terms for the title, author and/or publisher, together with a range of publication dates. For 
recent works, the ISBN can also be specified. 

It's worth bearing in mind that for older books the text that is being searched has been arrived at using 
O.C.R. (optical character recognition) software, and may be inaccurate if the quality of the scanned images 
is poor. It's also important to be aware that the details of author, title and date in Google's database are 
frequently inaccurate. In particular, for multi-volume works there is often no information about which 
volume is concerned, and the given date of publication may refer to the series as a whole. For medieval 
texts, Google usually credits the original author of the manuscript rather than the modern editor. 

Navigating
Once a book of interest has been found, Google allows the visitor to search for text within that book. The 
list of search results provides links to images of the pages where the search text has been found. 

Alternatively, the reader can browse through the book, viewing images of the pages. Normally (but not 
always), links are provided to the title page, the table of contents and the index. Once a page has been 
selected, above and below the image are displayed back and forward arrows to move to the previous and 
next pages, and an input box where a page number can be entered. The box normally accepts arabic 
numerals (1,2,3, ...) for pages in the main body of the book, and roman numerals (i,ii,iii, ...) for the title 
page, table of contents, preface and so on. The page image of the table of contents may also contain 
clickable links leading to the first pages of the sections. 

For out-of-copyright works, the reader can also usually download the whole book as a single PDF file. 

Problems accessing material outside the U.S.A.
As far as I understand, visitors from the U.S.A. can view all books published in the U.S.A. before 1923 in 
their entirety. In other countries, where copyright is determined by the length of time since the death of the 
author, Google is applying a much more stringent criterion based on the date of publication. For example, 
for visitors in the U.K. only works published before 1865 are available. 

To complicate matters further, Google also appears to be restricting all access to books published outside 
the U.S.A. later than 1908. 

Of course, the great majority of books published between 1865 and 1923 are legally out of copyright in the 
U.K. and most other countries. I have tried to clarify Google's policy on future access to these books, 
without success. 

A practical solution to the problem is to use an anonymous proxy service, which makes it impossible for 
Google to determine which country the visitor is in. One such service is Proxify, which provides a free 
service supported by advertisements, and also a faster advertisement-free subscription-based service. 
(Currently the free service is sometimes restricted owing to excessive demand.) The visitor simply copies 
and pastes the Google URL into a form on the Proxify website, and from then on can follow links normally, 
and even edit the Google URL to navigate to specific pages. 

Unfortunately it doesn't appear to be possible to download the PDF files containing whole books when 
using Proxify. In some cases the PDF files can alternatively be downloaded from the Text Archive section 
of the Internet Archive.